The issue of securities pricing for purposes of portfolio valuation is now front-page news (see last Friday's Wall Street Journal, for starters), seemingly every day. And with good reason: it gets to the very root of the trust relationship that (should) exist between portfolio managers and their investors, as well as the integrity of financial reporting.
Oh, yes, and the implications for investors are literally in the trillions of dollars, when one takes into account management and performance fees (hedge funds and long-only funds) and stock market valuation (banks, investment banks, insurance companies and publicly-traded asset managers).
With all that is at stake, wouldn't you think that the framework used to address this issue would be pretty objective and well-known by all? Well, if you thought that, you'd be pretty disappointed. The procedures for marking books are almost as varied today as when I started in the derivatives business in the early 1990s, when one could (and did) get incredibly creative when it came to marking incredibly complex books of correlation risk that lacked any kind of liquidity.
Bottom line: the issue is the same today as it has been for decades, with little in the way of improvement. Why? Because it has to do with one thing and one thing only: P&L and compensation. Because this is what makes the markets go around.
So if you'd indulge me for just a minute, I'd like to step away from the issues of P&L and compensation and to try and think about this issue rationally and pragmatically, and from the perspective of an investor. As an investor, what do I really care about?
- That the financial statements reflect both value and intention;
- That the financial statements truly represent Management's view of the business; and
- That, where possible, market-based pricing is used for financial reporting purposes.
As a pragmatist and as an investor, I know that market-based prices for all securities are sometimes either impossible or not reflective of true prices if my manager had to move the entire position. So assuming that I am using market prices in circumstances where they are available, what about the rest of my positions? I'd say that I'd want to know if the positions are held for trading (short-term) or investment (long-term). This gets to the issue of intention.
If positions are held for trading, meaning that short-term assets are being funded with short-term liabilities, then you've got to use either market prices or prices privately received from, say, five dealers, who are quoting based upon taking the bid (or at least the mid) side of the trade. And these are the prices that should be used for both NAV and performance fee calculations for funds, and carrying values for banks and other kinds of asset managers.
Let me repeat: If the asset is a trading asset funded with short-term trading liabilities, then you need true marks. No marking to model. Period. Because as we all know, models don't begin to reflect the realities of financial distress, and are inevitably skewed in favor of the manager, if not intentionally then at least subliminally because managers, by definition, tend to love their positions.
Conversely, if positions are held for investment, it must be demonstrated that such investments are funded with liabilities of like or longer duration. This way, an investor can take comfort in knowing that while the values used might not be market-based, the manager can ride out adverse market conditions and not be forced to liquidate at the worst time. However, investment assets should not attract performance compensation until they are sold, and Management must provide clear documentation as to the process used to value these assets for NAV calculation purposes.
This necessarily sets a higher return threshold for investment assets relative to trading assets, as should be the case: If one is giving up liquidity and the ability to collect quarterly performance compensation, then the expected return on these assets better be huge. This is where Management's view comes into play. This approach treats investment assets as if they were private equity in nature, being funded with long-term liabilities and attracting performance fees only when sold.
Therefore, a well-balanced manager would likely have a book largely made up of trading assets, with a lesser amount of extremely high-conviction investment assets that represent call options in a book of less volatile, lower expected-value constituents. So I am advocating a hybrid hedge fund/private equity valuation model to address the issues that have dominated the headlines since the subprime crisis broke out, but have really existed for decades.
From my perspective liquidity has always been the litmus test for managers; if you want to have a bunch of illiquid assets, then you should be able to support them with appropriate duration financing and defer performance fee gratification until performance has been proven. Otherwise, get those marks, guys. And if they suck, which they invariably will, too bad. You can't have everything. But investors need some stability, reason and fairness in financial reporting. I believe that if my approach had been used starting early this year, we might have seen problems emerging far earlier then we did.