It's the Liquidity, Stupid

At this point, the Bear Stearns subprime-mortgage-hedge-fund-meltdown has almost reached mythical iPhone and Blackstone status in my mind: I'm sick of reading about it. That said, there needs to be a little clarification of what the learning from the Bear Stearns situation should be. Is it the evils of derivatives? No. That subprime mortgages are bad? No. The dangers of leverage? Possibly. But there is a point - a single point - that links these and all the other explanations I've read together but that has been lost amidst the scramble to fill column inches and generate ad sales:

That liquidity, together with security selection, are the two most important variables impacting investment returns.

And in the case of many subprime mortgage securities in general and a goodly chunk of the Bear Stearns portfolios in particular, this is a key determinant of why they're getting wrecked relative to other professional investors that had positions in similar securities. If the liquidity profile of your portfolio is poor, your margin for error is very, very small. And liquidity needs to be analyzed via stress testing and scenario analysis, because a simple "there is a 1 point wide market today; these securities are pretty liquid" doesn't begin the capture the dynamics of a market shock similar to what we've encountered at the lower-rated end of the subprime mortgage spectrum.

I think it is critical to note that liquidity as a variable is closely related to risk management as a practice, which in the case of poor liqudity is severely hampered and has been on display in examples ranging from LTCM to Granite, Amaranth to Bear Stearns. So, if we are really going to focus on "root cause," IMHO the true root cause of why the Bear Stearns meltdown occurred is because of poor security and portfolio liquidity characteristics; all the other factors being thrown around are of secondary or tertiary importance.

This is also why Bear Stearns is delaying the release of a portfolio NAV until mid-July. From today's Wall Street Journal:

Investors in two Bear Stearns hedge funds will have to wait until as late as July 16 to learn how much money they have lost.

The Wall Street firm has had difficulty calculating the funds' fair value, apparently because many of the mortgage-related securities they hold are thinly traded and the market for them has been volatile.

Right. Poor liquidity, plain and simple.

Today's "Ahead of the Tape" in the Wall Street Journal titled Subprime Flu Sheds A Light on Derivatives is another example of missing the forest for the trees:

The subprime-mortgage crackup is casting a bright light on an often dark corner of Wall Street: derivatives.

********************

There are derivatives linked to interest rates, inflation and the weather. There are even derivatives of derivatives.

A rising concern is that many derivatives are "illiquid," or don't trade very often, making it hard to value them accurately. This can pose a problem for hedge funds, which generally need to place a value on their holdings every few months or so.

This is called being economical with the facts. First, I think the "dark corner" being referred to is pretty illuminated to be honest. Credit derivatives, structured notes, equity tranches of CMOs, CBOs, CLOs, etc. have garnered lots of column inches in the WSJ, the FT and a wide swath of the financial press. I don't think they're really such a mystery any more. And the statement that "many derivatives are illiquid," well, most are not. And the truth is that many publicly-traded common stocks are illiquid, so having a portfolio made up of illiquid equities containing no derivatives is toxic as well. So the logic here is flawed: the issue isn't derivatives, per se, it's liquidity, and liquidity is a variable in any investment. So let's focus on root cause, ok?

To be fair, the piece goes on to discuss a very important point that I've raised countless times on this blog, that from a NAV perspective, illiquid assets often get repriced infrequently, possibly leading to gaping disparities between carrying value and sale value:

The value might not always be accurate. Think of it like the housing market, where valuation can be difficult. If each house on a street of identical houses is worth $200,000, but then one sells for $150,000, you can bet that when the neighbors talk about the price of their homes, the valuations will be closer to $200,000.

The same is true for hedge-fund managers, except that they are playing with investors' money. A recent study by Paris risk-management firm Riskdata shows that roughly 30% of hedge funds that invest in illiquid securities smooth out returns with price estimates for these securities that are potentially self-serving, compared with just 3% for funds that invest in highly liquid securities such as stocks. The implication is that these funds aren't using market prices to adjust holdings' values.

In other words, a number of hedge funds have been tinkering with results, making them look less volatile. If and when these funds are ultimately forced to put an accurate price on the holdings, the outcome could be messy.

Now, maybe the author is over-dramatizing the point, but it is an important point and one that needs to be considered. The "stickiness" of illiquid asset values can lead to an overstatement of NAV (which leads to payment of higher management and performance fees) and the phenomenon called autocorrelation, neither of which is good. Now this is a point that warrants some focus and attention, particularly on the part of investors. Because investors in these funds aren't dummies and need to do their homework. There is no free lunch, and Mr. Market will be sure to penalize those who think there is. Bottom line, if you don't have the stomach for the volatility associated with large illiquid asset positions, stay away from funds whose documents permit it. Because all is rosy when things are calm, but when it gets stormy, look out.

Comments

[Yaser Anwar]

The actual volatility of a hedge fund can be much larger than its historical volatility because managers who trade illiquid assets tend to receive disproportionately large allocations according to mean variance theory.

Since the assets do not trade much, portfolio returns tend to have a large serial correlation on a MoM basis. The infrequent pricing = HFs may appear to demonstrate a smoother return curve. This lowers the calculated historical volatility.

It's dangerous to characterise a hedge fund purely by its historical mean, variance and correlation with other funds. Unfortunately, many FoFs already use mean variance optimisers to make asset allocation decisions and are reluctant to think outside the box for alternatives.

This is a good paper by MIT on 'An Econometric Model of Serial Correlation and Illiquidity In Hedge Fund Returns' for those interested in further reading.
http://web.mit.edu/mgetman/www/Files/Getmansky2003.pdf

[Espen Robak]

"Now, maybe the author is over-dramatizing the point..."

FWIW, I don't think he is. Volatile assets (that rarely trade) can get reported asset values hugely out of whack with their true "fair value" very quickly. Also, even though the investors in these funds are "sophisticated", they still receive audited statements in the mail saying these values are "fair value" - that's a lawsuit waiting to happen, for sure.

Consider, also, illiquid assets for which trading prices ARE available on a daily basis. I'm referring to restricted securities issued in PIPEs. This a huge market (DTCC claims 1.2 trillion in restricted securities outstanding) and there are probably 200 hedge funds that invest in PIPEs, minimum.

The issuers of these securities are tiny little microcaps (or what do we call the segment at the very bottom of the microcap market: nanocaps?) - so they're HUGELY volatile.

Also, the market evidence is that buyers of restricted securities want steeper haircuts when the volatility increases or at lower stock prices. Most HF managers, on the other hand, use some kind of internally generated "model" to value their restricted shares, warrants, or convertibles. (Think 30% across-the-board haircuts or something like that). This automatically dampens vol#s and - if prices head down - badly overvalues the portfolios.

[Jestyn]

Is there a business opportunity here? A sort of SGS (http://www.sgs.com/) for the financial markets providing an impartial pricing service...