March 08 CP.book Page 1 Friday, February 1, 2008 12:31 PM

The Liquidity Conundrum Paul A. McCulley Managing Director Pacific Investment Management Company LLC Newport Beach, California

For many years, economists believed that liquidity was a phenomenon that could be explained by a consideration of the monetary aggregates. But if the problems of the subprime mortgage market teach us anything, it should be that liquidity (or lack thereof) has much more to do with the appetite for risk on the part of borrowers and lenders.

I

n this presentation, I will talk about the nature of liquidity and put it in the context of what is going on in the current marketplace. My simple thesis is that liquidity is a state of mind. In the current market and talk of a liquidity crunch, investors are asking, where did the liquidity go? This question reflects the concept taught in basic economics that the money stock is the liquidity in the market, which creates the concept of liquidity as a pool of money. It is a very bank-centric concept of liquidity; the U.S. Federal Reserve (Fed) injects and withdraws liquidity, and the banks transform liquidity into deposits and loans. Thus, investors think of liquidity as a pool of money, and during a downturn, they wonder where the liquidity went. My answer is that liquidity is not a fixed pool of money but, rather, a state of mind and, in particular, a state of mind regarding risk. Liquidity is the result of the appetite of investors to underwrite risk and the appetite of savers to provide leverage to investors who want to underwrite risk. The greater the risk appetite, the greater the liquidity, and vice versa. Put another way, liquidity is the joining or separating of two states of mind—a leveraged investor who wants to underwrite risk and an unleveraged saver who does not want to take risk and who is the source of liquidity to the levered investor. The alignment or misalignment of the two investors determines the abundance or shortage of liquidity.

This presentation comes from the New Frontiers in Institutional Asset Management conference held in Sacramento, California, on 18 October 2007.

Liquidity and Nonbanks The current market reflects the differences between banks and nonbanks, or what I have named the “shadow banking system.” The shadow banking system includes hedge funds, conduits, structured investments, REITs, collateralized loan obligations (CLOs), collateralized debt obligations (CDOs), and so on. Each item is a levered investment vehicle. In that sense, they are similar to banks. But traditional banks are also very different. The primary difference is that banks are regulated. The regulators implement capital requirements that limit the amount of leverage a bank can have on its balance sheet. In return for accepting regulation, banks get two major benefits—deposit insurance and access to the Fed’s discount window, which translates into always having liquidity. Depositors generally do not care what banks are doing with their money because, in their minds, they are not giving their money to the bank but to the Federal Deposit Insurance Corporation (FDIC). They have no incentive to pull their liquidity because they are not really lending to the banker but to a third-party insurer—the U.S. taxpayer. If depositors decide to withdraw their liquidity, regulated banks can take the assets that they were holding on leverage down to the Fed, rediscount them, and maintain liquidity. That is the old-fashioned banking system. But traditional banks have not been the marginal source of liquidity growth in the last several years. Marginal liquidity growth has come from the shadow banking system—the group of levered intermediaries who take a lot of risk and are not regulated. As a levered intermediary without a regulator, they

©2008 Pacific Investment Management Company LLC. Reprinted with permission. All rights reserved.

MARCH 2008 ! 1

March 08 CP.book Page 2 Friday, February 1, 2008 12:31 PM

CFA Institute Conference Proceedings Quarterly

do not have a defined capital requirement; they can have as little capital as the open market will tolerate, and they can leverage more of their capital. But they do not have the two benefits of traditional banks— deposit insurance and access to the Fed’s discount window. Therefore, nonbanks are more susceptible to changes in the risk appetite of their deposit base, or funding sources that provide the leverage. The two primary funding sources for the shadow banking system are reverse repo and asset-backed commercial paper. When the market’s risk appetite is strong, the liabilities of the shadow banking system look stable. Brokers mark to market investors’ collateral at face value, require reasonable margins, and never hassle the nonbanks for more collateral, and the assetbacked commercial paper is picking up 2–3 bps versus conventional commercial paper. Everything looks fine. Every 45–90 days, brokers roll over the asset-backed commercial paper, and effectively, the shadow banks have the same stability in their liabilities as the traditional banking system. The only quasi-regulators that the shadow banking system has are such ratings agencies as Moody’s Investors Service, Standard & Poor’s, and Fitch Ratings. Before February 2007, the shadow banking system had high ratings from these ratings agencies and was easily funding itself with excessive leverage ratios. Nonbanks were able to hold an asset at a tighter spread than the regular banking system because leverage allows a nonbank to own an asset at a tighter spread relative to its funding cost and still generate an acceptable return on equity. For example, if one broker has levered 12 times and another has levered 50 times and they have the same return on equity objective, then the broker who is levered 50 times will buy all the assets. As remarkable as it seems now, this was reality at the beginning of 2007. Those who questioned the reasonableness of the spreads were told that it was all a result of an enormous pool of liquidity. In fact, that pool of liquidity was a union of risk-seeking states of mind between the nonbanks and their providers of liquidity. Supported by the ratings agencies, the reverse repo brokers and, even more importantly, the asset-backed commercial paper market were willing to take the offerings of more levered, less transparent, and more conflicted nonbank levered intermediaries. It should have been taken as a sign that the game was coming to an end when structured investments began to issue extendable asset-backed commercial paper, which gives the issuer the option to extend the maturity for a fixed time period. This effectively makes the buyer of that paper a lender of last resort, all for a mere 2 bps.

2 ! MARCH 2008

In February 2007, it was revealed that some of the assets that the shadow banks owned in the mortgage sector were not without problems. The marketplace did not pay particular attention to this situation until spring, when a Bear Stearns hedge fund revealed that its reverse repo lenders had asked for more collateral to ensure that they would not lose money on their investment. When Bear Stearns conceded that it did not have more collateral, the lenders decided to sell off the collateral that they did have. This was the wake-up call that initiated a run on the shadow banking system and the disappearance of liquidity.

Solving the Liquidity Conundrum All investment professionals have learned about bank runs and have concluded that traditional banks could not experience runs because deposit insurance effectively makes the government the lender of last resort. Shadow banks, however, did experience the modern day version of a bank run in 2007. The combination of not having the protections of the FDIC, being under pressure from reverse repo brokers to put up more collateral, and facing an assetbacked commercial paper market that refused to refinance took the market from seemingly unlimited liquidity to a liquidity crunch. The nonbanks could not sell assets fast enough to satisfy the increased risk aversion of their lenders. The liquidity dissipated and volatility returned to the market because investors’ state of mind changed their risk appetite. The issue of the liquidity conundrum is, therefore, resolved. My opinion is that there is no conundrum. A conundrum is something that cannot be figured out, whereas explaining the current state of liquidity is simple. Thirty years ago, people thought that explaining liquidity was simply a matter of figuring out where some component of the M2 money stock went. Today, determining the state of liquidity means looking at the state of the risk appetite of the levered-up nonbanks and the leveraged providers of liquidity to the levered-up nonbanks. The state of risk appetite is the critical pressure point in the market, and it hit that critical point late in 2007. In my opinion, the appetite for risk in the system can affect the decision-making process of the Fed because it influences the neutral real fed funds rate. The neutral real fed funds rate is not a static concept but, rather, a dynamic one based on the risk appetite in the shadow banking system. When the risk appetite is rising, the neutral level for the fed funds rate should also rise. More risk appetite means a higher neutrality, and less risk appetite means a lower neutrality.

cfa pubs .org

March 08 CP.book Page 3 Friday, February 1, 2008 12:31 PM

The Liquidity Conundrum

The Minsky Framework It is important to understand the phenomenon that I have been discussing in the context of the work of Hyman Minsky. Minsky, who passed away in 1996, was the leader of what is known as the “postKeynesian school of economic thought.” He spent most of his career at Washington University in St. Louis, but the last 10 years of his career were spent with a think tank in New York. His key contribution to economic thought was a critique of capitalism’s inherent boom/bust proclivities. He was a postKeynesian who was more willing than Keynes to take the logic of market-based capitalism in the financial markets to its logical conclusion. Minsky’s core thesis is known as the “financial instability hypothesis.” Translated very simply, the hypothesis states that stability is inherently destabilizing because stability leads to the extrapolation of stability into infinity, which encourages more riskseeking financial structures, particularly with debt. Therefore, the more stability a market has and the longer it lasts, the more unstable the foundation of the stability becomes. Stability is destabilizing because it begets more unstable debt structures. Minsky broke down the process of stability producing instability into three steps that are characterized by three types of debt units—hedge units, speculative units, and Ponzi units. In a financial cycle, a long period of stability leads to more marginal units of debt creation, and the economy shifts from hedge units to speculative units to Ponzi units. Once the economy reaches Ponzi units, it slows, and it is set up for a reverse Minsky journey. As I define the three units in the following, the process that led to the mortgage market crisis in 2007 will become clear. Hedge Unit. In Minsky’s framework, the hedge unit describes a borrower who obtains a loan to buy an asset, and the asset plus other income generates sufficient income to pay the interest and amortize the principal on the loan. The debt is self-liquidating, and it is hedged because the income stream can pay the interest and amortize the principal. It is a very stable unit, and in the mortgage market, a hedge unit would be a conventional 30-year fixed amortizing mortgage. In the past, debt was perceived as a bad thing, and therefore, trying to pay off one’s mortgage as quickly as possible was part of the culture. If the marginal debt creation in the economy is a hedge unit as described by Minsky, it is a stabilizing factor. Speculative Unit. A speculative unit is a step farther out on the risk spectrum. It is characterized by a borrower who buys an asset, and the income

cfa pubs .org

generated by the asset plus other income is sufficient to pay the interest on the note but not to amortize the principal. In the mortgage market, a speculative unit would be an interest-only loan with a balloon payment—at the set maturity date, a balloon payment is due that is equal to the amount originally borrowed. The speculative type of debt unit is less stabilizing than a hedge unit because the borrower is speculating on at least three things: The interest rate is not going to rise, the terms and conditions will not change, and the value of the collateral will not decline. If the marginal unit of debt creation is speculative, then the system is becoming less stable. But the paradox is that the longer an economy is stable, the more likely borrowers are to engage in such speculation. Doing so produces the immediately favorable effect of lowering the monthly payment because the principal is not being amortized. Ponzi Unit. Minsky’s third step is called the “Ponzi unit,” which is typified by a borrower who buys an asset, but the income generated by the asset plus other income is insufficient for amortizing the principal or even paying all the interest. In the mortgage market, a Ponzi unit would be a negative amortization loan—at the maturity date, the borrower has a balloon payment, but it is bigger than the original amount borrowed because of the unpaid interest. Like the speculative unit, the Ponzi unit is also speculating on the interest rate as well as the terms and conditions of the loan not changing. But it is taking a fundamentally different position with respect to the value of the collateral. In a Ponzi unit, the borrower is betting that the value of the collateral will go up. Borrowers who take on a Ponzi debt unit are betting that if they buy an overvalued asset, when the balloon payment comes due, another borrower will pay a higher overvalued price for the collateral. The collateral cannot just hold its value; it has to go up in value. Minsky and the U.S. Property Market. Minsky’s three steps precisely describe what unfolded in the U.S. property market over the last seven years. By 2006, the preponderance of debt creation at the margin was Ponzi unit finance. A classic example is the 2/28 subprime adjustable-rate mortgage in which borrowers put no money down, get a teaser rate for two years, and can opt to pay less than the full amount of interest. The borrowers choose how much interest to pay, and the rest is put toward principal. After two years, though, the interest rate goes up by about 500 bps. The majority of the marginal borrowers for the years 2004 through 2006 made use of this mortgage structure.

MARCH 2008 ! 3

March 08 CP.book Page 4 Friday, February 1, 2008 12:31 PM

CFA Institute Conference Proceedings Quarterly

Minsky’s hypothesis explains why the downturn occurred. Property prices had been rising at a steady and stable pace for a long period, so borrowers walked the path from hedge units to speculative units and, finally, to Ponzi units. In the midst of the exuberance, the rising prices were a self-fulfilling prophecy. As more people walked down the Minsky path, they drove up the value of the collateral. Very few defaults were occurring because borrowers do not default when they are making money. One should realize that what was happening, in effect, was that by 2006, the mortgage industry was granting to marginal borrowers a free at-the-money call option on the value of their property. As the property market continued to go up, the default rate on the mortgages was low because borrowers’ free at-the-money call options were going in the money. If they defaulted on their mortgage, they gave up the in-the-money portion. So, the default rate is initially low in the last stage of the Minsky journey from speculative to Ponzi. The rating agencies assumed that this default experience would continue. But by the first quarter of 2007, the subprime mortgages issued in 2006 had a surge of early payment defaults. The percentage

4 ! MARCH 2008

of borrowers not making the first payment on their mortgages rose quickly, which signaled that the property market had reached the Ponzi stage. For borrowers, the rationale behind not making the first mortgage payment can be explained by the call option effect. If the value of the property goes down, then borrowers’ call options are worth nothing, so why should the borrower continue to pay for it? Once affordability is stretched beyond any rational sense relative to rent values, borrowers stop seeking loans. The Minsky journey is over, and the economy starts heading in the other direction.

Conclusion The forward Minsky journey lifted the neutral real fed funds rate as the risk appetite of investors increased. But a reverse Minsky journey will lower the neutral fed funds rate and ultimately be stabilizing as Ponzi units are removed, speculative debt units are restrained, and hedge debt units return. The destination of a reverse Minsky journey is a more stable economy, but the journey itself will be a deflationary one. This article qualifies for 0.5 CE credits.

cfa pubs .org