2017 Financial Markets Conference—Policy Session 1: Shifting Sands of Low Interest Rates

Many central banks in the developed world dropped interest rates to the zero lower bound after the crises, and some have kept going into negative territory. But this creates problems as many institutional and individual investors require higher returns to meet their long-term commitments.


James Bullard: So I'm going to kick this off with what I'm calling "illustrative calculation of r†," it is illustrative, meaning that it's not as rigorous as I'd probably like it to be, but I think it'll hopefully get us off to a good start for the session.

So, you might ask yourself, "Why do I care about r†?"—which is the intercept term in the Taylor rule—and I'll just give you the lowdown in a few bullet points. First, unemployment gaps and inflation gaps are close to zero, the Taylor rule collapses to just a[n] intercept term plus the expected inflation rate. So what the Taylor-type rule recommends in this situation is you just set the policy rate i equal to what I'm calling r†, but the key question for policymakers is, "What the heck is the value of r†?"

We often call this a natural rate of interest, or natural real rate of interest, and I like to use the term r† to distinguish from r*, but it's certainly closely related—r*s become very closely associated with the new Keynesian model, but I'm talking about maybe a broader construct where we don't have to think in terms of that particular model. And you can divide it into three parts—and I knew you guys would really like some Greek letters going in this conference—so we've got λ, labor productivity growth rate; ψ, the labor force growth rate; and ξ, an investor desire for safe assets. So, basically what we're going to do is find out what all these factors are doing, and add them up and see where the r† comes to.

So if you had a strong desire for safe assets, then you'd have a large negative value for this ψ variable. I'm not using the new Keynesian model—I like to use the overlapping-generations model, for those of you that care about that kind of stuff—so there are demographics in there, and there are other factors. In this type of model, if you didn't have a desire for safe assets, this r† would just be equal to the output growth rate, that's one of the reasons I like it. It would certainly be [a] closely related concept in other models, but it would actually be equal to the output growth rate along the balance growth path, and that's one concept of the natural rate of interest.

So what the theory is telling you is that if you look at this in the data, this thing is supposed to have a constant mean. It should fluctuate around, but it should have a nice average value. So we're going to look at some data from 1984 to the present—I think the older data is of a different, higher-volatility era. So we'll start in 1984. We're going to construct an ex-post measure of r† by subtracting the Dallas Fed trimmed-mean PCE [personal consumption expenditures] inflation rate from the one-year Treasury rate. This will give you a one-year ex-post real rate of return in the U.S. economy—a safe rate of return. They're going to show a clear downward trend, macroeconomic theory does not like the downward trend. It hates the downward trend, it wants a constant mean.

So here's the picture: since 1984, this thing's going from about 500 basis points, all the way down to maybe -100 or -150 basis points. So it's a shocking, 600 basis point move over the 30 years—this picture's basically Bill Gross's career: "the greatest bond rally of all time."

But, like I say, the theory doesn't like this downward trend, so you've got to do something about it. You could stop here—you know, forget all the analysis I'm going to do—this is the problem right here. So, how are you going to explain this trend? You could just take this trend line and say, "OK, I'm just going to use that; I'm going to say that the r† is the last point on that dotted line there." That would be fine, too, but we're going to do more than that.

The declining trend does not seem to extend to ex-post real returns on capital; that return has actually been constant, so I think that that provides a reason to think that safe assets are an important component of this measure.

So here's the picture, the same data on the lower line, with the same trend, and then that upper line there with a trend through it is the ex-post real returns to capital. We've been staring at this picture at the St. Louis Fed, we think that this indicates that what has been going on has been increasing demand for government paper over this whole era, and so I think safe assets are an important part of the story. Not everybody thinks that, but I like that picture.

OK, so what's the main question? Which of the three factors is the most important in accounting for the downward trend: productivity growth, labor force growth, or desirability of safe assets? I'm going to use Markov switching, two-state Markov switching, so each of these factors we'll be able to take a high value and a low value, we'll try to fit that to the data in a rudimentary way. Those two values are called "regimes," and the idea is that these types of factors generally have constant means—something like population or labor force growth—but there can be infrequent shifts in the mean, and we want to characterize those shifts and see what that has to tell us about the future of r†.

We're going to do all three of them. So, first: let's look at labor productivity, that's been low. There are some papers by [James A.] Kahn and [Robert W.] Rich on the New York Fed website; they've got a low state of 126 basis points and a high state of 300 basis points. We were in the high-productivity growth regime from 1997 to 2004, according to this data. Here's a picture of the regimes: low regime from '84 to '97, high regime in the middle, low regime since 2004 or so, and actually the very latest data is mostly below the dotted line at the end of the chart there, so you could even argue for even lower productivity growth there.

Now, in labor force growth, that had been growing at 133 basis points annual rate until the Great Recession; since then it's been only 45 basis points. It looks like we're in the low-growth state, but if you squint at the picture I'm going to show you, you might argue that maybe we're making a return to the high-growth state, and so we'll consider both possibilities.

So here's the picture on labor force growth, '84 to the present. You've had a pretty constant mean up until the crisis, since then it's been very low—and so you have a lower mean, only 45 basis points—but if you look at the data toward the end of the chart here, some of that is actually above the previous mean in the previous regime. Maybe we're going back to a higher growth rate for labor force growth. Of course, there would be a lot of papers you could consult about this kind of issue, but statistically speaking, possibly you're going back to the high regime.

What about investor desire for safe assets? We're going to treat this as a residual here, so I'm going to take the regime switching trends for labor productivity and labor force growth out of the raw data on ex-post safe real returns, and that will leave us with a time series. That time series also has a declining trend, and I'm going to fit two-state regime switching to that and interpret that (the two states) as a strong desire for safe assets versus a more normal desire for safe assets.

So here are the two regimes you get from that: basically in the earlier era, a kind of normal desire for safe assets, and in the later era, a stronger desire for safe assets. This is adjusted data, now, that takes out the labor productivity and the demographics.

The estimated value for the ψ are -304 basis points in the high desire for safe assets regime, and a +63 basis points in the normal desire regime, so the difference between these is gigantic, a whopping 367 basis points. So this factor is the most important factor of these three—other things happen with productivity and with demographics, but this one would be much larger.

Here's the probabilities associated with the high and the low come on the model.

So what does this all imply for the natural real rate of interest? Well, here's a chart that shows what we just got, the high state and the low state for each one of these factors; if we were in the high state for all three factors, the r†—which is at the bottom of the first column there—would be almost 5 percent. These things have occurred in the data before, so you could have an r† as high as 496 basis points, according to this, and if everything was in the low state, which is one of the arguments we're making here, then it's -133 basis points, and the difference between those is over 600 basis points. So this is a huge difference, so that what's the advantage of going in this direction? You can assess whether you think these three factors are moving into their high state or are going to stay in their low state, or what is going to happen, and that's going to affect your ideas about what's going to happen to r† in the future.

We've got labor productivity in the low state—so we'll set that to 126 basis points; ψ, the demographics in the low state—45 basis points, and high desire...but, it could be the ψ is switching to the high state, which would be 133 basis points, or also appears to be a higher desire for safe assets—that subtracts 3 percent.  So, r† is either -133 or -45, depending on how you view the labor force growth; then if you add the inflation target back into this, kind of a back of the envelope calculation, you'll get a Taylor-type rule policy setting of either 67 basis points or 155 basis points today. That would be a neutral setting for the policy rate, given that the gaps are zero. The actual policy rate is 88 basis points, which is bracketed by those two values, and so you could argue whatever you want about the labor force growth part.

So the policy rate would be approximately at an appropriate setting today, according to this analysis.

Now, you've read the literature, I know you guys pay a lot of attention to this. It's a large literature and it's rapidly growing, there are a lot of papers, they tend to be a lot more sophisticated than what we've got here. The only point of this is just to think in terms of regime switching, which is an alternative way to characterize the stochastics for two of the three factors: the labor productivity and the desire for safe assets. It definitely appears that we're in the low state, and it definitely seems like we're not moving to any high state based on what's in the data. So this suggests that the natural rate of interest—and hence the Fed's policy rate—can remain low over the forecast horizon.

So I guess what I'm objecting to in some of the other analysis is, they'll estimate a low value but then they have only a single mean, and so they'll always have the r* or the r† returning to the mean; whereas, if you've got the regime switching, you would not predict necessarily a return to anything because it depends how these factors move around between their different regimes.

[Thomas] Laubach and [John C.] Williams is the most famous paper on this; [Vasco] Curdia is a similar analysis, it gets a low value for r*. This paper by [Marco] Del Negro and coauthors has a demand for safe assets and comes to a similar conclusion that I've come to here, and so you can check out those papers. The analysis will provide some background on how you might think about recent trends on the natural safe rate of interest in this regime-switching context. The natural rate, according to this, is very low—likely to stay low—and so the policy rate is likely to also stay low over the forecast horizon. If you want something more detailed and rigorous and with a structural interpretation, you can go to the paper by Del Negro, et al., which just came out in Brookings (or is coming out in Brookings).

So, thanks very much for your attention, and I appreciate your...I hope it was a good start to the session, and I'm going to turn it over to Trish to continue the session here.

Patricia Mosser: I'm Trish Mosser, I'm the director of an initiative on central banking and financial policy at the School of International and Public Affairs at Columbia. My comments today are going to focus more on the financial stability and macro stability implications of very low interest rates, across the globe, particularly the advanced economies, as well as the U.S.

So, to get one question of the way quickly, that was asked of us: Are they here to stay? Yes. Extraordinary monetary policy may be responsible for getting us to the really exceptionally low—and in some countries negative—policy rates, as well as very low long-term rates, due to forward guidance and asset purchases, but at this stage, low-productivity potential growth seems likely to be with us for quite some time, and as a result the equilibrium real rate is likely to remain low for all the...and in addition, for the reasons that Jim just said.

And by the way, that's no surprise. If you look across decades, across countries, across financial crises, financial panics not only create very large recessions and depressions that are hard to get out of, but they lower long-run growth rates—significantly. So I guess if it's any comfort, there's nothing unusual going on during this period, but what it means, as Jim so clearly pointed out, is that when monetary policy normalizes, what it's normalizing to is not that much higher than where rates are, at least in the United States—other countries have a lot further to go.

One policy aside, not about monetary policy, of course, structural changes—and in those countries that have fiscal space, fiscal policy that was oriented more toward long-run growth—might of course boost some of the numbers in Jim's little model, but those don't seem to be forthcoming, and they're certainly not happening where it seems to be needed the most, which I would argue is in Europe and Japan.

So, monetary policy it is. Let me focus on the risk, both the macro and financial stability risks, around monetary policy, in a very "low for long" world, and a key question is, what's the bigger risk, is it higher rates, or lower rates? And to think about that I want to talk for just a couple minutes, as background about the monetary policy transmission mechanism and how it is working and has worked in a really low-rate environment.

How does monetary policy work through the financial system to impact the real economy? Let me start with the obvious here, and that is that central banks, when they buy long-term assets or set their policy rates very low or negative, directly increase the market value of existing assets. That means their capital gains, their increases in financial wealth, their increased collateral values for investors, and all of those things can give a boost to economic activity, and in fact, there's a fair amount of evidence since the crisis that they've done exactly that, and that's been one of the major transmission channels.

It's also clear that the impact on market interest rates, of negative rates and asset purchases and so on, has been immediate and large, and that, of course, is likely to work in reverse, if not more than 100 percent, on the way out once the rates are increased.

But what about the impact on credit and banking channels of monetary policy? In a very market-based financial system like the United States, the impact of the credit channel of monetary policy seems to have worked...I don't want to say quite "normally," but it's worked, meaning that easy money—low rates, asset purchase programs, and so forth—has spurred growth in largely market-based credit in response to the policy changes. And when it's reversed, the odds are that we'll have a reverse of some of that as well, as we typically do when policy rates are normalized.

But in bank-based systems—and here I'm talking particularly in Europe and Japan, where rates are very low, negative—asset purchases are still going on and are likely to continue to do so for some time. This story's different—and I believe it's particularly different for those that have implemented negative policy rates.

First of all, in those economies, the bank lending channel—not the market credit channel, but the bank lending channel—is very important, and secondly, by reducing the long-run return on assets—by lowering long-term yields or setting negative policy rates—that has some potentially counterproductive side effects (that would be the nice way to put it) on the ability of banks and other leveraged financial institutions to actually lend more.

Why? Well, they're basically four things that are interacting with each other. Banks have, A) huge reserve balances at central banks, a lot of which are subject to negative interest rates, they have lower returns on new lending—now, they do have the capital gains I talked about a minute ago, but all their new lending is at lower rates, they have very limited ability to reduce the cost of their liabilities, which are largely in the form of retail deposits where the zero lower bound—not the effective lower bound, but the zero lower bound—binds. To my knowledge, there's not a single economy with a negative policy rate right now that actually has any retail deposit rates below zero; they have not moved, but the returns on all the assets have gone down.

And let's be frank, in many of these cases, a not insignificant share of the banking system is capital constrained as well. That's not a prescription for a lot of credit growth.

So the risks of lower rates here may be high. And by the way, central banks really are well aware of this—another aside—and the problem with the bank lending channel, and in fact, it's reflected in the way they've actually implemented negative policy rates. If you look at the way it's been done, it's basically to shield the banks and, to a certain degree, depositors from the side effects; first of all, they haven't pushed the rates very negative. Secondly, only a fraction of reserves—and in some countries it's a very small fraction of reserves—actually has to pay the negative interest rate or is subject to negative interest rates.

So the marginal cost of reserves, if you like, is negative, but the average cost is a significantly higher, or closer to zero, number. But because the marginal cost of reserves is negative, and that's the main policy tool, that means that the policy transmission to market interest rates—wholesale funding and sovereign rates—is working pretty normally.

This is a picture from last summer, but I picked it because it was sort of the nadir, the low point, of negative rates—but that's an impressive chart. The whole red area is the sovereign yield curve out to 10 years; the red area is negative and the blue area is positive. The numbers aren't quite as negative now, in fact, because growth has picked up in a number of these economies, so I'm not saying negative rates don't work, I'm simply saying they have side effects. It's not quite as much, but it's an awful, awful, awful lot of the sovereign debt around the world that is subject to negative interest rates. It's down from its high of 11 1/2, but I still think it's about $8 1/2 trillion.

I also want to reinforce a point that [keynote speaker] Raghu Rajan made last night: because the policy transmission mechanism through market interest rates has basically worked very effectively, that means the policy transmission mechanism through exchange rates has worked really well as well—in fact, maybe the most effective channel of monetary policy in many of the economies that have negative rates. I think that's correct, and I think it should lead everyone to be concerned if the credit channel is only partly working, but the exchange rate channel works really well, where are you really getting your bang for the buck?

It's probably not so much in credit, and this is a picture of credit to GDP in the euro area—and, remember, in the euro area, growth has been exceptionally low, [and] credit growth has been even lower. They're back to 2004 levels of credit to GDP.

Now, part of that deleveraging was clearly necessary, but a lot of this period—the last few years—has a very stimulative monetary policy, and it has had very limited effects on actual credit growth.

So what does this mean for financial stability, particularly if we believe that low rates are here for a while? Well, first, continued low profitability of financial investments in banking and so forth. So, does that mean there is going to be more risk taking or less risk taking?

Let me take the "more risk taking" side of that. Will investors take "too much risk" (if we could define that clearly)? Simple measures certainly tell you that they have taken more: credit risks are incredibly narrow, term premia are actually negative, the percent of long-duration bonds that's been issued—I think the duration of the U.S. bond market has gone from 4 1/2 to 6 over the course of a few years, [with] the federal government being part of a big chunk of that. And credit spreads are very narrow, on top of that, spreads are narrow for much lower-quality credits with very few credit protections.

So, yes. When rates go up, does that mean those assets are going to lose money, that investors are going to lose money? Almost certainly. Will defaults go up? Probably. But the financial stability question is different than that. The question there is, has the greater risk taking been very large, has it been highly levered and widely held credit instruments that have somehow been transformed into money-like liabilities with a lot of leverage? If the answer to that question is yes, then the entire system has a problem.

So far—so far—that sort of really excessive overleveraging and risk taking does not seem to have happened...except in a few areas, perhaps. I'm pulling out now—there are lots of indicators that we could look at—I'm pulling up a summary indicator from my former colleagues at the OFR [Office of Financial Research], which happens to be their financial stability monitor, and without going through all the details, red is bad and green is OK is the general color scheme here (so everybody knows).

So there's lots and lots of interest rate risk, everybody's way over their maturity benchmarks. Lots of demands for long-term safe assets in particular—some in positioning, some in credit risk—reflecting particularly what's going on in corporate lending and some commercial real estate lending. But, honestly, there's a lot of green and yellow there—and in particular, the contagion channels that we tend to think of (which is a lot of maturity transformation, and a lot of leverage—at least, to the best that we can measure it) seem...not so high. Certainly nothing like one would have seen in, say, 2006.

So I'm a little less concerned about this—not concerned that people won't lose money, because they will—but more concerned about the stability of the system.

On the other hand, if you're more concerned about the problem I was talking about a minute ago—which is about less credit formation, with low returns and the potential drag of low rates—then the place you might want to focus really is on the biggest global financial institutions, because they have all the things that I just listed, plus (obviously) a higher capital bogey and liquidity bogey that they have to hit over the next few years.

Now, on one hand, higher capital and liquidity ratios for really large, globally systemic banks is sort of the whole policy design, right? It was partly to tackle "too big to fail" and have them internalize the externalities they posed to the system, so having those firms shrink, and do less credit lending, and pull back out of certain businesses and markets, was sort of the regulatory intent. And that's kind of what they've done, honestly, there are plenty of examples globally of the financial activity shift that's actually going on. For example, the very largest global SIFIs [systemically important financial institutions] have given up market shares in businesses in big parts of the world to other smaller regional and smaller international banks, and frankly, that's probably a good thing, from a diversification standpoint.

Moreover, in theory it should encourage non-banks, who don't have to hold, first of all, all those negative rate reserves—and probably have different liquidity and capital rules—to move in—aggressively, even—and take on more credit risk from banks. And has that happened? A bit. One of the problems is, here, we don't measure this worth a darn, because we don't have very good data. But are there examples? Certainly. Private equity, fintech firms, asset management—even insurance companies—are expanding those sorts of business.

Is that good—for near-term growth? Almost certainly. For long-term financial stability? Maybe. One can make the case that if the shift in activity is all about regulatory arbitrage, however, and water flowing downhill (where requirements are the least), then have we just moved a set of risks from (partly through regulatory reform) our relatively strong hands into weaker hands with more leverage, less cushion, less liquidity, and more fragility?

I don't have the answer to that question. I wish I did, but I think it's an important one and a key one to understand what the risks really are.

Let me very briefly mention two more things—two macro risks, as I call them. One is very much a global risk, and it's related to something that Raghu Rajan said last night. What happens when liquidity risk really starts to shrink—when central banks stop doing the sort of maturity transformation intermediation that they have been doing? This is the global liquidity risk question, and it's a key, particularly for the benchmark currencies, and it's really particularly an issue, I think, for the U.S. dollar.

As the BIS [Bank for International Settlements] has pointed out in a series of papers, borrowing and lending in U.S. dollars, in markets and instruments, and by financial institutions that have nothing to do with the United States—corporate bonds, lending for international trade and so on, infrastructure, even real estate—all those things globally depend on U.S. dollar liquidity outside the United States.

Right now, I want to point out that the world is awash in that, in dollar liquidity—has been for years. And what happens when first the Fed's, and then later other central bank balance sheets start to shrink? The possibility of a risk of a dollar shortage outside the United States seems very significant to me. It circles back to the question of how big does the Fed's balance sheet actually need to be? I don't think anyone has the answer to that question today, but it's a risk that everybody should be keeping an eye on.

And finally, of course, what I call the "big macro risk" is basically the lack of monetary policy space on the downside. Balance sheets are huge, equilibrium real rates are low, if there is another really big macro shock, how much bandwidth do central banks really have to respond?

And on that cheery note, I'll sit down.

Scott Sleyster: I think this is the point where we move to the other side of the stage, and you're talking to an insurance company and a pension plan—the people that have to live with the implications of a sustained low rate—and as you'll see shortly, we tend to think we are in a low-rate regime. I think "regime" is an interesting way to look at it.

So what I'd like to do is spend the next 11 minutes talking about how we manage our insurance portfolios in our segments, in a manner that's highly defensive really for any kind of environment, but for this environment, so I don't want to stand here and say low rates don't cause pressure—they do—but I think in the context of the way life insurance companies manage their business, I think you'll find it's more like a headwind than actually a very severe threat.

So let me talk about how we run our portfolios in life businesses, and we're very strong and disciplined with this at Prudential, but quite frankly, when I spend time with my contemporaries at other large life companies, we generally run our businesses the same way, so I don't think it's unique.

When we invest in our portfolios, we start with the liability. My job is to make sure I construct a portfolio that, if you will, allows me to meet the obligations—the commitments—that I've made to my customer, and I'm trying to preserve my pricing economics or my return on capital throughout the life of that product.

Life insurance companies generally are not in the maturity transformation business, we're not taking short-term deposits. In fact, we get long-term deposits...many of those deposits don't actually have prepayment flexibility or optionality built into them, and that allows you to construct a portfolio largely out of fixed-income instruments.

And then we tend to run with a pretty high-quality portfolio, and that's driven by a number of factors. One, we're in the business of making long-term promises, so people are going to look for an AA rating, for example, as a benchmark. When we're competing for a pension liability and people are using advisers, we have to meet the safest available standard from the Department of Labor.

So we have market pressures, obviously, we're highly regulated—we are a SIFI, and a GSII [globally systemically important insurer]—we also have a lot of state regulation in insurance. And, you know, the capital standards penalize after a certain point of equity and alternative risk. It gets pretty punitive, and so that's going to be a pretty small part of your portfolio.

So I thought it might be best just to simply work through an example—and this is one that's gotten a lot of attention, and we've gotten a lot of questions about the pension risk transfer business. When a large corporation wants to lay off...they've decided not to offer a traditional pension plan anymore, they want to lay that off on someone else, so that they don't have to spend so much time on their earnings calls talking about whether they're a health care company, a pension company, or an auto company or phone company, or something like that.

So this would be a classic example: over the last five years we've done almost $50 billion in pension risk transfer transactions in the U.S. The average age of the participants in our plan I think has typically been about 71, 72 years old. What that means—and what this chart is—is the amount of payments that we've signed up to make to participants in retirement, and what you see that is obviously, if the average age is 72 by the time you hit the 30-year mark...most people won't be around at 100—and you see that little gap over on the right is what we would call the tail liability: the liabilities that we expect to run beyond 30 years.

So in effect, I have the ability to go into a high-quality bond market, and defease, more or less, not technically defease, but match off against that liability, and pretty much hedge that risk away and lock in my profit margins. Now obviously, we'll have credit cycles, and there'll be some cash flow variability—people won't exactly die on schedule—but it's pretty darn predictable when you're working with large pools. OK?

Also, it's important to note that pension plan participants don't have any optionality, they can't turn in their pension, there's not a gray market for pensions, so they really don't have much of a choice. Actually, we give them an incentive to live as long as they can.

So here's what you end up with, if you will, from a cash flow sensitivity, and this is actually how we look at our liabilities. In our $400 billion portfolio of product liabilities, we've chopped it up into segments around products, and so we actually have over 150 different segments where we have somebody that's portfolio managing that product. And what we end up with—and this is a sample pension case—you look at the interest rate sensitivity that you have. So the red line or the red bars on that chart are showing you where we would want to line up to perfectly match that liability. So hopefully, one of the questions you're asking now is, "Well, gee, why aren't you perfectly matched?"

Well, perfection's a hard goal. But I'd say more importantly, Prudential tends to do a lot of lending in the private market—so commercial mortgages and direct lending to corporations in a private placement market, if you will, Reg D-type products—those products are really terrific for us. One, it allows us to diversify away from strictly public names—which, when you get a large portfolio that gets to be a challenge for everyone—but, more important, in those markets we have covenants that are very valuable when you get into that tough part of the credit cycle.

But probably more important, in essentially all of those securities, we have some form of economic call protection, so that if the people want to prepay they have to make us whole, which allows us to go reinvest, but it gives us more premium to protect our returns and preserve our promises and our margins.

So what happens when you have a product like this? You then add it into a portfolio with lots of other products, you then look across that legal entity in net, and then when you end up with net gaps—and it's never going to be perfect—then you'll go in with derivatives—right?—to do an overlay, to make sure you're comfortable, that you're in your risk tolerances.

So, one, that's why you would expect to see some level of interest rate derivatives in an insurance company. Under Dodd-Frank, we have to have more postable collateral available, and so you've actually since seen an increase for insurance companies that have to use derivatives, moving into some lower-risk assets, postable collateral—also, moving toward exchanges. OK?

So, in this environment—and given the fact that we're maintaining a relatively tight duration discipline on assets supporting our insurance liabilities throughout that 30-year period—the impact of low rates is relatively modest. Look, I don't want to say it's not a headwind—it is—and that chart I just showed you was the assets supporting the insurance liabilities. What it didn't show is the capital that we have—we have capital standing behind those products as well—and then we have tail liabilities that run beyond 30 years. They're not so much on these pension products, but on life products you often have cash flows out in the 30-, even 40- or 45-year period, depending on when people are buying life insurance.

But, since for the bulk of your products, and the ones we're selling right now, you're in that investable horizon, you've done a pretty good job of protecting yourself from interest rate movements of any types. Patricia made the comment that if rates change, you're going to be experiencing a loss, while on a mark-to-market basis, and depending on how your accounting works, you might be having a mark-to-market loss—but if those participants don't have a right to prepay and my bond prices go up or down (but, quite frankly, I'm holding them to maturity), I'm really well matched, I'm not really suffering a loss, what I've really done is preserve my original economics. And if you used the same discount rates on the assets and liabilities when you were marking them to market, then I wouldn't really be experiencing much of a loss.

The last point I have here is more of a statutory insurance comment, but if we get it wrong, if you will, and we have mismatches or we have a really severe credit loss and we take some bond losses, every year our actuaries are going in and testing various interest rate regimes, and they're looking at whether or not we have the capital to meet our obligations. And if we don't, we have to strengthen reserves, and any of you that have followed insurance companies, you'll know that at certain points and times, and on certain products, you'll get something wrong and then then we have to pony up reserves. OK?

So let me speak to, then, two more topics that come out of that. One: one of the questions in the pre-call was, "What do you think about low rates?" This is an illustration used by one of our economists, and I would say basically, Prudential has basically bought into [the idea] that we're going to be into a low-rate regime for a pretty long time—certainly not with the degree of sophistication that President Bullard just used, but our view was taking an extremely long time horizon, and when we didn't have a period of extreme high inflation, risk-free rates kind of bounced around 2 to 4 percent on the 10-year, and that I think kind of makes sense, if people aren't looking for much of a real return on particularly safe short-term assets.

And we think what happened is, with the inflation and breaking the back of it, we have been in an extremely long bond rally and people have been expecting us to go back to a new normal, if you will—but not an old normal. I think the challenge for us as an insurance company is to make sure our pricing actuaries weren't looking just back 20 years or 30 years, but in fact were taking a very long view. Because some of our products do have recurring premium, and you have to make an assumption about where forward interest rates will be; they're not all pension risk transfers.

So that's generally been our house view, and then the last thing I think I would respond to is, I think a lot of people assume, then, that because we've been in this low-rate environment, there's been a real reach for yield in the insurance companies or other portfolios. I can't speak to other portfolios, but what I put together here was a chart using total life—so I think that has some P&C [property and casualty] in there, but that was the best data I could get for what I was looking for. But I have Prudential's asset mix on the right, and what you see is that at the bottom of this chart, going into 2006, we had almost 14 percent of our portfolio in higher-risk assets—so that would be the below-investment-grade bonds and equities, and alternatives—today that's actually down to 9 percent, and that's kind of been our new normal. If you look at the government securities up at the top, you can see that we went from 4 to 8 percent—that's because we have to hold more collateral, and in some cases when we want duration, we don't want to add to corporate credit so we'll just go out long on Treasury securities.

And on the left side, I have the global industry asset mix, which I think has been relatively constant. So, look, I don't want to tell you in a low-rate regime people aren't going to reach a little bit for yield, I think they will. I think there'll be a tendency to hold a little more yield-y assets in the alternative space, but because of our rating aspirations, because of the regulatory regime we're in, and because of the disciplines that we apply on our portfolio with matching, I just don't think you really see too much of that.

So why don't I stop there, and we'll take questions later.

Satish Swamy: So today, I am representing the University of California Office of the CIO, and I'm going to talk about what are some of the challenges we are seeing, and how we are coping up with this interest rate regime change that President Bullard referred to.

Now, one thing, keep in mind, as an investment professional, there is never really a dull moment in the investment business. Every day's a new day, every day you have news, for example, you have the French elections right now, we're analyzing the Brexit recently, the U.S. elections, geopolitics, uptick in terrorism, North Korea, China. There's a lot of uncertainty, and how these uncertainties impact investments, so that's what we are tackling. And if you want to throw in, outside of this uncertainty, the discussion about if fiscal stimulus is going to potentially disappoint, and GDP maybe is going to not budge from here, then are we going to be in this low-rate, low-return environment for a long time?

So there's so much uncertainty in the system. Amidst all this uncertainty, ladies and gentlemen, there are two things that are certain. Number one: as a pension manager, I have mandated pension obligations to meet. As an endowment manager, I have a mandated endowment payout policy that I have to meet.

So, what is really the scope of the problem I'm dealing with? This is the scope of the problem, so, if you look at this slide... Basically, at the University of California [UC] we manage four different products: the endowment portfolio comprises the endowments of the campuses, of the labs, the hospitals across the UC system; the pension, or the defined benefit plan, for the UC employees; the retirement savings, or the defined contribution [DC] plan, where we give the employees of the University of California a plethora of investment choices for them in their DC plan; and finally, the working capital pool, comprised of the total return in the short term, which is basically cash belonging to the different campuses in the University of California system that is left with us in the investment office to invest these assets and to generate returns.

So this is really the scope of the problem we have to deal with on a daily basis: four different products. In these four different products, how have we allocated at the entity level? This is what the allocation is at the entity level: public equities is 50 percent of the portfolio, then you have fixed income, and then you have the other investments, you have some cash sitting, and that's really the asset allocation for the entity right now.

In the interest of time, what I'm going to do is, I'm going to spend the next 10 minutes—I have exhausted my 2 1/2 minutes, so I have 10 minutes remaining—to explain to you our strategy in the defined benefit—or the pension—plan, because it is very difficult to go into the other products. And the pension actually puts into perspective a lot of the challenges and a lot of the evolution that has come about in the UC system.

As a pension manager, this chart is very simple but very profound, take a look at it. This is the monthly index returns going back to 1987, that has what the market has given to a portfolio manager in the pension space. The top line is the S&P 500 U.S. Market Index, you have the Bloomberg Barclays U.S. Aggregate Index on the fixed-income side, and then, if you want to throw in some diversification, you have the MSCI ACWI, or the All-Country World Index, ex USA. OK?

Now, what have these indices given the portfolio manager in terms of returns? So if you were invested in the S&P 500 since 1987, annually you have received about 10.4 percent, and in the bond space, you have got about 6.4 percent, and the MSCI ACWI has given you about 3.4 percent. OK?

Now, let's talk about asset allocation. If you were allocated 60 percent in equities and 40 percent in U.S. bonds—and this is strictly a U.S.-only portfolio, all right?—and you had zero allocation to international, it was pretty good—8.8 percent if you were just passive, and I'm not going to go into a discussion about active management because a lot of the financial press has been talking about active versus passive, how active managers have underperformed. What is given to us is the index, and let's say you're invested in the index, and you got 8.8 percent—not bad! And if your hurdle rate, your benchmark on the pension side, was about 7 and change, you beat your hurdle rates, all right?

Now, let's say for kicks, actually, you introduced some diversity into your portfolio, and you threw in the MCSI ACWI ex USA—which, as I said earlier, was not very attractive—but if you throw in the rate of return, the annual rate of return over the last 30 years was not bad. It was still, you know, had a 7 handle, 7.4 percent.

At the University of California, we did a lot better than the first allocation of 60/40. How? Basically, the investment staff in the late '70s and the early '80s believed in three things: number one: [they] had immense confidence in the Fed to fight inflation—and, as a result, loaded up on long duration bonds. And you saw some charts here, some fascinating charts here, about yields, real rates, going back to the '70s and '80s. So [they] loaded up the portfolio in the late '70s and early '80s on long duration bonds, bought every zero-coupon bond they could get their hands on.

Second: [they] had immense confidence in the U.S. economy, invested in large-cap U.S. growth stocks. Now, remember, just to give you a little bit [of] history, when was the microprocessor really invented? In the '70s; in the mid-'70s, it came into production in the late '70s with the Intel 8085 microprocessor. So, the investment staff back then had the foresight to look ahead for the next 20 years and say, "You know what? The U.S. is going to be a great place to invest."

So [they] loaded up on large-cap U.S. growth stocks. And finally, [they] had a lot of confidence in the ability of the United States to innovate—the entrepreneurialism of the United States—and invested in private equity. We were one of the first pension plans to invest in private equity, and the only private equity we had back then was venture capital: California, Silicon Valley. We had excellent relationships with venture capital firms, and we had a stake in venture capital.

So those three asset classes didn't care much about asset allocation, only focused on speeds. It was the need to generate returns, we were return seeking, invested in the best returning asset. And the investment team back in the late '70s and the early '80s believed, had immense confidence in these three factors, and loaded up the portfolio in this.

What did the returns look like? The returns were fantastic. The returns for almost 20 years were in the high teens—easily beat your 40/60 combination of stocks/bonds. It was so good, ladies and gentlemen, that in 1989 the board decided to suspend both the employer and the employee contributions into the pension plan. Now, you show me if there is any pension plan out there with that kind of pedigree: for 24 years, neither the employer nor the employee contributed into the pension plan. Unbelievable, all the returns were generated from the assets.

But things changed, in the early 2000s—about 15 years ago—things changed. We had too much risk in the portfolio, somebody concluded. So what did we do? We hired a bunch of very smart consultants, who came in and they recommended that we be invested in assets—everything under the sun—and, hence, that asset allocation that I just showed you, all right? Till that time, everything was managed internally, we didn't have any external managers. At that point, we decided that except for fixed income (which is still managed internally), everything else was farmed out to external managers, and that's when we changed our allocation, we moved into more...the Lehman Agg, or the Bloomberg Barclays Agg-type, fixed-income benchmark, has a lot lower duration; moved into international equities; moved into emerging-market equities; moved into real estate, real assets—you name it—and hedge funds.

So we also introduced hedge funds as a strategy, but now, remember, that allocation still worked. It was not that great, compared to the previous allocation, it still worked because [of] the philosophy and the culture of the institution to focus on security selection. We picked the best managers, and we picked the best asset classes, and we allocated it, but we were constrained by asset allocation. We couldn't deviate from a given asset allocation. OK?

On the fixed-income side, also—just to give you a little bit of history, we were very cognizant about the "reaching for yield" pre-crisis. A lot of the smart people on Wall Street tried to sell us CDOs [collaterized debt obligations], CDO-squared; a lot of the rating agencies stamped a AAA rating to mezzanine tranches of subprime securities—I could never figure out how those were AAA. We never invested in any of those, because we didn't understand, we told all the smart people on Wall Street that we don't understand these cash flows. "In fixed income, if you want to invest, we want to first understand the cash flows, and we're sorry, it is too esoteric for us." And the rest is history, all right?

So, that's the allocation right now, and if it's done well, and we have reinstated contributions into the pension plan...and, what about now? So, I have just explained two different regimes: in the '80s and '90s, [and] the asset allocation right now. What about the next 20 years, how's it looking?

One of the key drivers, we feel, that has driven this low-rate, low-return environment is globalization in the last 20-plus years. Global supply chain and trade has created increased capacity, and even now, even today, we feel there is significant excess capacity in the system. This has driven—globally—has driven down secularly in both inflation and interest rates.

Monetary policy is also globally more synchronized, global rates, global GDP, are low. And add to this discussion what President Bullard said about, especially in this business cycle, falling productivity, we are concerned. And if you add all this that I just said into the mix, we are going to be lower for longer, in both rates and return. I don't want to stand here to discuss the real neutral policy rate, but if the real neutral policy rate was 2 percent pre-crisis, we feel as portfolio managers the real neutral policy rate is close to zero right now. OK?

So that's the ingenuous part, so what are we doing right now? So, how are we looking at things, and how are we going to adjust to this regime shift? I'm going to spend a couple of minutes to explain that the new thinking behind other, and we are setting ourselves for the next 15 to 20 years, we are once again going back to our roots and focusing on [the] best risk-adjusted returning asset[s].

We're going to be benchmark agnostic, especially on the fixed-income side, [we] don't care about benchmarks. The most popularly used benchmark—the Agg, the U.S. Aggregate—has a lot of interest rate risk, and we just heard Trish Mosser say that there is a lot of interest rate risk in the system. Oh, yeah, there is a lot of interest rate risk in the system. OK? We are going to lose a lot of money if interest rates go up, so what do we do? We are going to once again go back and think about how we are going to allocate—in the fixed-income space—how are we going to allocate things differently? We still need a fixed-income portfolio security selection. We are not going to focus on returns of the index, we are going to focus on the hurdle rates, and the hurdle rates argue that there is going to be, tactically, a lower allocation to fixed income over the next few years.

We also strongly believe that there is no specialization in the investment business anymore. Every investment professional has to talk fixed income, equity, international equities, absolute return, private equity, real estate. Everybody should be talking about investments across the space, because you cannot be siloed anymore, you have to go after the best returning asset.

And also, in terms of collaboration, we feel now that with globalization there is going to be a lot of opportunities globally, so you need to be able to pick reliable, trustworthy partners that have alignment of interest globally. You need to work with partners, you need to work with other large asset managers. Asset classes all compete for capital, you cannot pigeonhole and say, "I'm going to have 30 percent in bonds, and 70 percent in equities"—can't do that anymore. Every asset class competes for capital, you have to be invested in the best returning assets.

And finally, we are going to go back to security selection, and move away from asset allocation. We at the University of California feel that we would like to be at the forefront of this change, and what we have done at the University of California could potentially be a classic business school case study; OK? But here's the problem, implementation takes time and patience, especially with pension funds, it is not that easy, implementation. To change the mind-set of the board, to change people to think differently, takes time, and we believe that time and scale is our pedigree, we have the patience to do that.

Ladies and gentlemen, thank you very much for your time. That's my spiel, and we're going to stick to it. Thank you.

Bullard: Thanks very much. Welcome to the question and answer part of this panel, you're supposed to be diligently writing down your questions or voting for other questions.

And I've got some here, and one that is...of course, these are all moving around on me now, so I can't...there we are. OK. One that people are asking that I think we could all reflect on has to do with safe assets, so I think what I'll do is I'll give a pitch about the two stories that I've heard about this question, and then I'll let you guys...because I would like to get feedback on these stories, so...

What I said was that desirability of safe assets has gone up over the last 30 years, and this is having a profound impact on the general level of rates. There are two stories, one is a global story and the other is a regulatory story. The global story goes like this: you divide the world in half, and half of the world issues government debt and the other half of the world—the other half of world GDP—for various reasons, they don't issue very much government debt, or no government debt at all. And this government debt is useful in financial markets because it has some liquidity value, it has value as collateral, and for other purposes.

So, because half the world's not issuing, the other half has to issue, in some sense, twice as much, and they're not doing that. So, as the emerging markets are growing over time, you're not getting enough safe assets issued to really take care of all of world GDP, and so what's happening is that the prices of these things are skyrocketing, and the yields are going way down.

So that's one story, that's a global story. The other story that I hear in conferences like this one is a regulatory story—it could be related, but it's a regulatory story where over the years regulators have piled up requirements for different types of pools of managers of assets, that they have to hold a certain percent in certain types of assets, or have capital requirements—what counts as capital? Well, how about sovereign debt? And so you get more and more of this over time—and some of that might be even informal, where you have boards of trustees saying, "I'm uncomfortable going more than such and such allocation outside of what are considered safe assets."

And because of that you get this buildup in demand over time, and there just isn't anywhere near enough issuance to take care of that kind of demand, and so what you have is skyrocketing prices and declining yield.

So those are the two stories I know, and I have not seen them fleshed out in enough detail that I'm really convinced by either one, but I don't know if the other people on the panel have any reaction to those two.

Mosser: I do think the global story has merit. In fact, I'd add a wrinkle to that related to something that I said during my comments, which is that the demand not just for safe assets outside of the United States—and, by the way, saving—and it relates back to my comment about benchmark currencies and the role of the dollar...the dollar is, and all liquidity is, what makes global financial markets work, let's [put the] cards on the table here. And as a result, when emerging-market economies feel that for macro reasons that they want to do a very large chunk of their government savings, and the savings in their economies—at high cost, by the way, to their economies—in the form of foreign assets, because they feel they have to do it, being...as I say, the economies that are at the end of the dog's tail, the rest of the global economy moves and the end of the tail does this [waves arm around like a dog's tail].And they feel they have to provide their own insurance—macroeconomic insurance—and predominately those are going to be in dollars, and they're going to be in safe assets.

And so, in addition to that, I think because global finance and global trade, and a whole bunch of other international businesses such as commodities and infrastructure, all run in dollars as well, and all the borrowing and lending associated with that is all in dollars. Short-term, safe dollar liquidity is particularly at a premium, and there's been lots of that, and if there's not a lot of it, then you go buy a long-term asset (or an intermediary buys a long-term, relatively safe asset), and then does maturity transformation on that, it gives you short-term dollars—so you can go run your other businesses all around the globe, that may not have much to do with the U.S. business place.

So I think it sort of feeds upon itself. My view, at least, is that certainly regulation, and liquidity regulation—particularly for the biggest global players, that have to hold more, do more self-insurance themselves—has also increased demand. I don't know if that's alone enough to explain what's going on, I doubt it, frankly, but certainly that has to play a role as well.

So, I agree with you, with a few wrinkles around the edges.

Sleyster: I think I'll be brief on this, but you saw in our pattern, largely for the U.S., that we've taken up our safe assets probably by 50 to 60 percent—a small percent of our portfolio, but proportionally a very big step because of the Dodd-Frank collateral posting-type rules. We really don't have a high need for liquidity in a life insurance portfolio, so we didn't need it for liquidity purposes.

The one observation I guess I would make is that we're also an international insurance company, and I think actually tying in to your remarks, Patricia, about [how] the rest of the world is heavily banked, and the capital markets are not nearly as well developed. And so, in an awful lot of the countries—even a big country like Japan—the actual bond market, if you will, or kind of a private placement market—or even, quite frankly, a commercial market—is so heavily banked that it doesn't really...one, you don't get paid that much for nongovernment assets because they are so competed for, so you have to do a risk-adjusted analysis and say, "Is an eighth or a quarter worth the credit risk I'm taking on?"

And so in those portfolios, we'll continue to hold a lot of government securities—although we will tend to hold long-duration government securities.

Swamy: So, from the University of California perspective, there's really nothing safe about any asset. I mean, if you're investing in only safe assets, then you're definitely not meeting your benchmark—your actuarial rate of return. So we have a small percentage of our portfolio in safe, liquid-type instruments. But you know, it's an active, defined benefit plan; we are a going concern, it's an active endowment pool. And in the endowment, actually, we have the luxury of investing for, really, the long term, and we have reduced our allocation to very safe assets. And in fixed income, it is not...and that's what I keep telling everybody on my team is in fixed income it is not the return on capital, it is a return of capital.

So, "How safe is your capital when you deploy?" is, "Are you going to get your money back?" So that's kind of the mind-set.

Bullard: OK, let's try this question for Scott: a lot of people think pension funds are underfunded; isn't this a major problem? We've got low rates, we're saying low rates are going to continue into the future. So what's your take on this?

Sleyster: Well, certainly the answer to that question is, in fact, "Yes." I think in the U.S., and globally, many pension schemes are underfunded. In fact, I think a pension's a little bit different than an open university endowment, where you're still keeping people in a plan and having those contributions, but for most of the plans in the U.S.—and quite frankly, it's pretty common in Europe—there's been a lot of plan terminations. And at that point it becomes less of an evergreen-type of a situation for investing in, and actually starts to become more like the pension obligation that I showed you up there.

So, yes, a lot of plans are underfunded, and under-gap the way their mark-to-market is primarily off of the discount rate based off of a bond portfolio. So, as rates have come down—but quite frankly, as spreads have come in—the underfunding challenge has gotten pretty big.

Two observations I guess I'll make from that: first of all, Prudential's pension plan—and I think most of our life peers, but certainly Prudential's plan—is substantially overfunded, and we haven't had to make contributions for a long time, and the reason it's substantially overfunded is because we applied a liability-driven strategy to that portfolio, once it no longer became evergreen and it became, in essence, a liability that you needed to hedge or defease, in one way or another. And then, of course, as rates have come down, we've had a lot of bond gains.

So that's the model that works. I think the average funding last year was 85 or 88 percent; markets have rallied since then, but it's probably not much more than 90 percent in the U.S. And usually when you close out a liability by bringing it into an insurance company that's going to take the risk down and hold capital, you'll end up having to close it out at 105 or 107, or something like that. When you really mark it to market, there's usually more that you have to put into it.

So companies have to fund that deficit; the PBGC [Pension Benefit Guaranty Corporation] fees are going up here in the U.S.—they're pretty punitive, particularly on small balances—so I think that's the challenge that in particular corporations have to deal with, they're going to have to pay for it slow over time, or if they happen to have a windfall gain from, say, a change in tax policy, an offshore money comes in, or something, they may decide to take a portion of their plan and defease it.

And that's really what we've laid off on an insurance company, that's what we've experienced. Typically, people don't come in and terminate their whole plan, they'll take retiree-only, or management, or what have you—they'll just take off a certain bite at a time.

Swamy: Actually, I would like to add, since we manage an active pension fund [at] the University of California, we have actually...it was a very attractive, very lucrative pension plan. And over the last two or three years, actually we have cut back, not only have we introduced contributions, but we have actually made it less lucrative, the vesting period has gone up, contributions have gone up, and new employees coming in have the option of selecting between pension plans. So that's one of the things we have changed.

Bullard: OK. Well, we're getting a lot of votes on this one, so I'll have to go with this one. This is about balance sheet reduction, so, there's been a lot of talk that the Fed would...the Fed has a $4.5 trillion balance sheet, and that we'd allow that balance sheet to start to run off—possibly later this year—so this question is, would this create a taper tantrum-type reaction?

So, I'll give you my take on the taper tantrum, and others can chime in. The taper tantrum occurred in 2013, but the actual taper began in December of 2013. So I just want to remind everybody of the sequence of events, because I think the taper tantrum was a communications issue and not a matter of actual policy.

What happened was, in May of 2013, and also at the June 2013 meeting, then-Chairman Bernanke made some comments that maybe the Fed would start to slow the pace of asset purchases; this was a surprise to markets, and it didn't go over very well. At the June 2013 meeting, the committee actually didn't do anything—didn't change the statement at all—but Chair Bernanke, during the press conference, talked about some scenarios under which the FOMC [Federal Open Market Committee] might start reducing the pace of purchases. Well, this led to about a 100 basis point global yield rise on longer-term yields, and that's what's known as the taper tantrum.

Now, in September of 2013, the committee actually didn't do anything, but because we didn't do anything and the expectation was that we were going to do something, we actually got an easier-than-expected reaction to that—and so, really [we] had these two key meetings where the committee didn't do anything, and then finally in December, we did begin the taper—and at that point there was actually no reaction in the financial markets, and the whole taper went off really without a hitch during 2014, and we stopped purchases in the fall of 2014.

So to me, the lesson from this is that this was a communications problem, and not a matter of the actual supply and demand dynamics in financial markets.

Mosser: I agree that it was very much an episode that pointed out the communication challenges of starting balance sheet reduction—and it's clear the FOMC has taken that to heart; I actually hosted the vice chair a couple of weeks ago, who spent the entire evening talking about exactly that issue—and so I do think it's a challenge, it's unique. The Fed hasn't done this before, it's not like raising the policy rate, which...I do think there is a risk even with the best communication strategy, of markets front-loading more of what I'll call the "term premia" reaction upfront—they certainly did it on the way in.

All of it, if it's communicated clearly and carefully, and the strategy for how the FOMC is going to proceed, I think that can be minimized. Would it completely eliminate it? My best guess is there may be some, but I don't know that that's completely unavoidable, markets are going to do present discounted values pretty well, and so it won't be surprising that they do some of it.

Is it a big risk? I think as long as the communication is clear, and it's not huge, and it is sort of the appropriate change in term and risk premia—it's one of those situations that I have to assume the Fed is thinking about and understands could happen.

Sleyster: I'll make two quick comments more, from a market participant perspective. If you take a buyer out of the market of that size, it's clearly going to have an impact, and I think the bid offer among folks seems to be, that's equal to one or two rate hikes. So I think the market participants believe it's a pretty significant event when it occurs, and so I think the idea of being careful and measured about it—which certainly seems to be the way folks are handling it thus far—makes a lot of sense.

But when you spend time in a trade room, it has a very different feel than this conference, or this room, it's a little more emotional, it's occupied by ADHD human beings who are watching screens, and you have the risk of a lot more reactions that perhaps aren't as well thought through. I think probably the biggest breakthroughs in finance in the last 20 or 30 years have been more around human behavior maybe than about anything else, and I think it's pretty hard to manage that, so there's always risk.

Swamy: So, from our perspective, basically, the balance sheet reduction comes under two things, right? The Treasury portfolio and the mortgage-backed securities portfolio. On the Treasury side, we feel that when maturities come to you, they will not let everything run off, we don't feel that's prudent.

So part of it, they will let it run off, and they will allocate the same percentage, even to mortgage-backed securities. Right now they're spending about 20 to 25 billion per month in buying agency pass-throughs—they will probably not buy 20 to 25, but at least 10 to 15, and slowly taper that—so we're not really that concerned, then.

And I think the Fed has done—to Tricia's point, and Scott's point—they've done a good job communicating to the markets, and I think there will be a "once bitten, twice shy"-type mentality, and they're going to be very cautious.

Bullard: OK, everyone wants to know, is there a bubble in equity markets? So, I'll let whoever wants to start with that chime in.

Swamy: So, is there a bubble? Bubbles come, bubbles go, but as a portfolio manager, we have to be invested. As I alluded to earlier, security selection is very important, valuations are looking rich right now, but valuations have been looking rich for a while now. OK? So if you thought valuations were looking rich three years ago, it is looking extremely rich right now.

So we are cautious, you know, we are more nimble now, we have cash in the portfolio. And if markets gyrate and there's a knee-jerk reaction, we are willing to provide liquidity to the marketplace—and probably buy, because we have faith that the economy is not going to go into recession any time soon. And we have faith in global growth, and the central bank's ability to control inflation, and maybe keep things in check right now.

Sleyster: We don't own a lot of equity, our combined equity/real estate/hedge fund/private equity is only 4 percent of the portfolio. It's a big portfolio, so the numbers are relatively big, but it's not a big percent. "Bubble" is a blinking term, it's a power term. I think we'd say valuations look at the rich end—cap rates are very low on real estate, equity forward multiples are close to pre-crisis levels—so there's a lot of indicators, but we don't know when it ends, and these things can run for a long time.

So, in general, what we've been doing is maintaining our portfolio allocation, not letting the runup in valuations push us from four to five or what have you, and so we've been harvesting to the extent we can stay within the range of our targets, but we haven't been relaxing our targets and letting it run.

Mosser: So all of this, as viewed through a financial stability lens, which, I guess the short answer is, no, although equity values are rich by almost anybody's calculation, I certainly agree with both the gentlemen to my right. Is it the sort of worry that, when it corrects itself—even if it's a pretty abrupt correction—that it's going to have very large-scale spillovers and big impacts everywhere else? No, a lot of people are going to lose money, but not of the sort of "financial system with lots of contagion" sort of worry.

Bullard: I would agree with that. I think the assessments that we've done in St. Louis and elsewhere, that the...try to get a sense of the financial fragility of the system, I think we are...while there are pockets of vulnerability, I think overall we're not in anything like what we saw during the housing bubble or during the tech bubble in the '90s. So, [it] certainly looks like equity valuations are at the high side of historical experience, and we'll just have to see how that evolves going forward.

So, I think we are out of time, so that'll have to be the end of this session. Thanks for your great questions, and thank you for the really excellent presentations. I really appreciate it.