An Investor's Guide to Hedge Fund Leverage (Part 1)

Leverage (more specifically deleveraging) difficulties have caused more than their fair share of pain of late in the hedge fund world.

We're going to explore the concept of leverage by way of an example for now; let's suppose:

1. we're in a simple world (let's call it Wonderland) in which the price of a bond or a loan remains constant at $100 per $100 of par;
2. managers can only buy (go long) assets, and no naked shorting is allowed;
3. AA bonds carry coupons of LIBOR + 90 basis points (i.e., L + 0.9%); and
4. A bonds carry coupons of LIBOR + 125 bps (i.e., L + 1.25%).

Imagine these bonds are fairly illiquid and so the manager's strategy is to buy them and hold them to maturity. Assuming zero defaults, the manager who buys AA bonds will earn L + 90 bps on his/her investment, while the manager who took on more risk with single A bonds will earn L + 125bps.

Let's put aside our feelings about whether LIBOR + 1% or so is a "good" return. One thing we know for certain: unless LIBOR goes crazy and hits the 29% mark, your hedge fund is never going to earn 30% annually by simply buying and holding bonds, even if there are never any defaults. We're not going to invest in a hedge fund that targets L + 1%; to compensate for the limited upside potential of bonds/loans, relative to equity, fixed-income hedge fund managers need a little more zip. To "hit" the big numbers (assuming no shorting for now), they need leverage.

Leverage can be "achieved" in various forms, such as total return swaps and repurchase agreements ("repos"); for now let's examine the essence of the mechanism.

Suppose hedge fund HARRY has $10mm of capital and wishes to buy a $10mm position in that AA bond paying L + 90 bps. HARRY could buy it, after which HARRY is fully invested and yielding L+ 0.90%.

Alternatively HARRY enters an agreement with bank that wants to sell HARRY the bond. The agreement says that bank will LEND HARRY $10mm to buy the bond, subject to the following conditions:

1. HARRY must post the purchased bond to the bank as collateral for the loan;
2. HARRY must pay bank LIBOR + 10 bps on the loan, as long as it's outstanding; and
3. HARRY must post 5% haircut against the AA-rated collateral (to mitigate the bank's risk that the collateral defaults or depreciates in value).

Summary: HARRY's collects L+90 on this bond, pays L+10 to the bank, and has to put down $500K capital (5%) as haircut. Thus, he captures L+ 90- (L+10) = 80 bps on this transaction (of size $10mm). Since he's posting 5% haircut, he can logistically (we're still in Wonderland) perform this operation 20 times (20 x $500K = $10mm = HARRY's total capital). In so doing, he's making:

20 x 80bps x $10mm = a handy 16% return p.a. on capital

Some market terminology for you:

• HARRY is 20 times levered;
• HARRY manages $10mm capital, but has $ 200mm of assets under management (AUM);
• The bank's lending rate to the hedge fund is often termed the "Pricing Rate"; and
• The "haircut" described may also be referred to as initial margin, but bear with me here - we haven't yet breached the topic of "variation margin."

The better the collateral quality, the lower the haircut. Using a 1% haircut in the above example, would allow HARRY to lever up 100 times in Wonderland. Municipal bonds -- historically typically AAA rated and often insured -- are an example of collateral that could be highly leveraged on the back of this strong credit quality.

Now let's have a look at some actual (slightly historical) numbers, while remaining for a while in Wonderland...

(Click on the table to enlarge it)

This concludes Part 1. Part 2 will follow shortly, and will discuss, among other things:

• deleveraging, and its effects
• variation margin - and what happens with 2008 numbers
• leverage facilities: diversification, termination
• and possibly, the cases of Bear Stearns Asset Management (and Merrill Lynch), and the municipal arbitrage deleveraging nightmare