Leverage can be dangerous. There is no shortage of hedge funds that have collapsed as a result of excessive leverage, with Long-Term Capital Management, detailed in Chapter 13, being the classic example. Investors have learned the lesson that leverage is dangerous rather than leverage can be dangerous. In this sense, they are much like the cat that sits on a hot stove. As Mark Twain observed, “She will never sit down on a hot stove-lid again—and that is well; but also she will never sit down on a cold one any more.”
Investors always seem to ask hedge funds the question: “How much leverage do you use?” This question is flawed on two fundamental grounds. First, the question is meaningless, given that it ignores units of measurement: the underlying investment (that is, what is being leveraged). Second, it implicitly assumes that there is a direct connection between leverage and risk. Not only is this assumption false, but it is even possible—in fact, entirely common—for a higher-leverage investment to have lower risk.
Consider a comparison between two fixed income funds that are approximately equivalent in terms of credit risk, liquidity, and other relevant risk factors with the exception of exposure to interest rate changes. Assume Fund A’s portfolio is unleveraged and has a net duration of 10 years, while Fund B’s portfolio is leveraged 5:1 and has a net duration of one year. (Duration is the approximate multiple by which a bond’s price will change, in percentage terms, given a small change in interest rates. So a seven-year-duration bond would decrease approximately 0.07 percent in response to a 0.01 percent increase in interest rates.) An investor comparing these two funds on the basis of leverage would conclude that Portfolio B was five times as risky. In fact, even at 5:1 leverage, Portfolio B is only about half as risky as Portfolio A because its holdings (on an equal dollar-invested basis) are only one-tenth as risky as Portfolio A’s holdings.
Leverage can even be used as a tool to lower risk. As a simple example, consider a long-only fund that implements 2:1 leverage and uses all the leverage to fund short positions. In this instance, although the fund has gone from unleveraged to 2:1 leverage, it has transformed the portfolio from 100 percent long to market neutral in the process, clearly reducing risk.
In both these cases, focusing on leverage alone leads to totally erroneous risk perceptions because the holdings in the portfolios of the funds being compared are vastly different in terms of their risk. Risk is a function of both the risk of the underlying investment and leverage. Comparing investments only on the basis of leverage without considering the underlying investments leads to nonsensical risk-comparison conclusions. In essence, it makes sense to compare leverage only if the underlying investments are approximately equivalent. The underlying investment is, in effect, the unit of measurement.
When I worked for a fund of funds group, as part of our monitoring procedures we set up fund-specific flags related to such factors as gross and net exposures, size of monthly loss, significant changes in assets under management, and leverage. These flags were used to signal when we should take a closer look at a fund. One month, the leverage flag for a credit fund in our portfolio was triggered. The threshold level for the leverage flag had been set to 5:1 for this fund. When we examined the situation, we found that while leverage had increased to 5.2:1, its highest level ever, the leverage increase was entirely due to a substantial increase in the fund’s short hedge. In fact, the gross short exposure was 1.7:1, while the gross long exposure was 3.5:1, resulting in a net long exposure of only 1.8:1, the lowest level in two years. Moreover, the beta of the fund versus the high yield index was only 0.6, meaning that the fund would be expected to decline by only a little more than half the decline in an unleveraged high yield bond portfolio (as implied by the index).
This real-world example, which represents a common situation, reflects two ironies.
This actual example illustrates an important principle: If leverage is used for hedging, it actually reduces risk rather than increases it—exactly the opposite effect of leverage assumed by conventional wisdom.
Imagine a teller at a bank being told by her supervisor to count up the money in the till at the end of each day, but to calculate the sum based on the number of bills without regard to their denomination. Sound ridiculous? Well, that’s exactly what investors do when they attempt to gauge risk in terms of leverage without considering the risk of the underlying portfolio.
Some investors operate with checklists of investment criteria. Frequently, one item on the checklist is the maximum permissible leverage level. For example, a pension fund might have a rule that says it will invest only in hedge funds that use no more than two times leverage. Such a rule might sound like a rational risk constraint and one particularly appropriate for a risk-averse institution such as a pension fund. Yet such a one-size-fits-all rule represents the height of folly, leaving the institution open to investing in highly risky funds that use less than the maximum leverage or even no leverage at all (e.g., long-only emerging market fund, long-only technology fund, etc.), while avoiding many low-risk funds (e.g., market neutral). A maximum leverage constraint applied uniformly to all prospective investments regardless of portfolio content is analogous to a traffic law that applies a 40 miles per hour speed limit to all roads, in all conditions. Such a speed limit would be both absurdly low for a highway in fair weather, but recklessly fast on an ice-covered, curvy mountain road. Certainly, it would be much safer to drive 65 miles per hour on the former and 15 miles per hour on the latter than 40 miles per hour on both.
In essence, leverage comparisons should be made only between equivalent portfolios. A corollary is that asking for the average leverage of the underlying assets in a diversified fund of funds portfolio makes no sense. How can you average leverage across completely different investments? Assume a simplified example of a fund of funds that has three holdings: a fixed income fund that uses 6:1 leverage, a market neutral fund that uses 2:1 leverage, and a long-only fund that doesn’t use any leverage, all with equal allocations. What is the leverage of the portfolio? This is a nonsensical question that can’t be answered any more than the question “What is the sum of two apples plus five baseballs?”
Imagine you receive a check for $1,000,000 from a very wealthy anonymous benefactor (just as in The Millionaire TV series of the late 1950s) with the one twist that the gift is conditional on it being invested in one of the two following investments for exactly one year. Which would you choose?
Investment A
Investment B
Compare the alternatives and pick the preferred investment before reading on.
Now assume the same scenario, but this time you are given a choice between the following two investments. Which would you choose?
Investment A
Investment B
Again, compare the alternatives and pick the preferred investment before reading on.
If you are like most people, you will have readily picked investment A in the first case and investment B in the second. But here is the odd thing: In both cases, both the return and worst loss of investment B are twice as high as those of investment A, or equivalently, investment B is twice as leveraged as investment A. Why then do most people have a strong preference for A in the first case (i.e., 25/–25 combination preferred over 50/–50) and the more leveraged B alternative in the second (i.e., 20/–4 combination over 10/–2), even though both involve the same leverage (2:1)?
The explanation for this apparent paradox is that people’s desire for a more or less leveraged version of an existing strategy is influenced by the following two factors:
The higher the return/risk ratio and the lower the risk level, the more likely leverage will be preferred by an increasing portion of investors.
There are certain circumstances when leverage can be especially dangerous. Leverage can imply high risk when any of the three following conditions apply:
Although leverage can be dangerous, the knee-jerk reaction many investors have to leverage can lead to nonsensical investment biases. Investors need to focus on risk, not leverage. Sometimes, leverage may indeed be a risk factor. But other times, such as when it is used for hedging, leverage can actually be a risk-reducing factor. In this sense, blanket prohibitions against leverage are shortsighted and misguided, as is the use of leverage as a risk measure—that is, the assumption that higher leverage implies greater risk.
Leverage is a tool that can aid more efficient investing. It allows for creating hedged portfolios with higher return/risk ratios than unhedged portfolios and for adjusting the return levels of low-risk investments to more closely match investor preferences. But like any tool, leverage can cause damage if misused. The solution is not to ban its use—any more than it would make sense to ban power tools because they are dangerous in the hands of intoxicated users—but rather to ensure that leverage is used appropriately. The guiding principle should be risk assessment, where risk is viewed comprehensively as a function of both the underlying investments and leverage, not naively as a function of leverage alone.