An Uneasy Look at Leverage

Reprinted by permsion from REIT Wrap Special Report, March 2, 2010

By David Shulman*


Every REIT CEO and CFO should tattoo on their foreheads the following inscription, “Mike Kirby is right.” Mike Kirby is, of course, Director of Research at Green Street Advisors. He has been a voice in the wilderness calling for REITs to deleverage their balance sheets on the easy to understand grounds that higher leverage is statistically associated with lower long term shareholder returns and that unleveraged commercial real estate, which declined by 30% (likely an under-estimate of the true decline) in the early 1990s and 40% from 2007-2009, simply cannot afford to have leverage ratio in excess of 50% of gross value. Two crashes in 20 years are a bit much for an “asset class” heralded by the pension consultants as “safe.” By the way my UCLA dissertation written in 1975 found no gains to leverage for the few REITS that were in existence from 1963-74.

Moreover there really isn’t any support in financial theory for REITs to be leveraged. Kirby rightly notes the famous leverage indifference theorem of Miller and Modigliani which states that in a world of no personal and corporate taxation there are no gains to be had by leveraging up the balance sheet and in fact there is the risk of the enterprise bearing the costs of bankruptcy. As General Growth is showing us, those costs can be quite substantial.

The real “advantage” from leverage comes from the tax deductibility of interest at the entity level which REITs don’t have. To be sure interest payments at the REIT level can shield income from taxation at the personal level when it is paid out as a dividend, but to the extent that REITs are owned by tax exempt institutions that advantage goes away. Put simply REITs at best receive minimal benefits from leverage, but all of its costs.

If all of what I am saying is true, why do REITs leverage up? My own answer is that leverage has been in the DNA of every real estate professional ever since the first developers in the ancient city of Ur came on the scene. Besides, leverage is made for long-lived assets with relatively stable income streams. That is why institutions usually like to finance income producing real estate, but just because there is a willing lender doesn’t necessarily mean there should is a willing borrower. Nevertheless there are willing borrowers because the empire builder that resides in every real estate professional cannot just say “No” to the crack dealers of Wall Street.

So if theory says “No”, and the evidence says “No”, but existing practice obviously says “Yes” to leverage, where do we end up? What the market might be saying, to state the obvious, is that low leverage is benign and high leverage is malignant. Thus the degree of leverage a REIT takes on should not jeopardize the firm under stress conditions. The questions are where do you draw the line and what is the right measure for leverage.

First we have to dismiss the most widely used measure of leverage, debt/gross asset value. Why? Debt/gross asset value is a poor measure because it is very cap rate dependent. For example, at the height of the recent boom many REITs thought themselves to be conservatively financed with a 60% debt ratio. However when cap rates moved from say 6% to 9% the debt ratio exploded to 90%. In order to avoid the spot cap rate problem a more appropriate ratio to use would be debt/EBITDA which takes into account the ability to service the debt including amortization independent of fluctuating cap rates. Furthermore this metric has the advantage of being widely used in the corporate bond market.

I would argue that 5X EBITDA is the appropriate leverage metric for most REITs. Indeed a ratio of 5X EBITDA gives a great deal of credit for the stability of real estate cash flows. Remember the typical investment grade industrial corporation normally does not exceed 3X EBITDA.

Let’s assume that a REIT owns 10 buildings of equal value and it historically trades at around 12X EBITDA. In the old days a prudent owner would have mortgaged seven of the buildings at 60% and kept three unencumbered for both safety and opportunistic purposes. In this example the leverage ratio for all 10 buildings would be 42% (70% X 60%) or 5X EBITDA (42% X 12).

I would, therefore, argue that 5X EBITDA is a normalized maximum level for leverage for most REITs. If you want to quibble, as long as apartments have access to lower rate Freddie and Fannie financing at 70% of value, you could certainly make a case for 6X EBITDA for apartment REITs. Conversely for the operationally cyclical hotel REITs, debt should not exceed 3X EBITDA. Needless to say that with most REITs trading in excess of 7X EBITDA, the deleveraging process that began last April still has a long way to go.

Where do pension funds come in? It is no secret that many large pension funds and endowments with leveraged real estate investments got clobbered in the recent downturn. In my view it was an inevitability brought on by a complete misperception of risk. Put bluntly leveraged real estate is different from unleveraged real estate. It is much riskier. The higher risk can be categorized in two ways.

First whenever a pension fund engages in a leveraged transaction, it is, in effect, shorting it own bond portfolio. How so? All a pension fund has to do to understand this concept is to “gross-up” its leveraged asset and fully consolidate it on its books. For example take a hypothetical pension fund with the following asset allocation: $600 million equities, $300 million bonds and $100 million leveraged real estate/private equity.

In this example a $100 million real estate/private equity investment with $300 million of debt on it would be booked as a $400 million asset with a concomitant $300 million liability. In reality by booking only the equity, the pension fund is under-stating its investment in real estate and over-stating its investment in fixed income securities as the real estate debt cancels out the bond assets. Distilled down the pension fund portfolio is holding $600 million in equities and $400 million in real estate/private equity -- not exactly what you would call a “conservative” asset allocation. When looked at this way is not a surprise that a seemingly small allocation to real estate had such a drastic effect on pension fund performance.

If you prefer a different approach, the second way to look at leveraged real estate in a portfolio is to view it as a warrant. People tend to forget that in the original options pricing paper by Black and Scholes common equity was viewed as an option to buy back the firm from its bondholders. When leverage is low this notion is of minimal significance, but when leverage is high it becomes very obvious and it is this insight that theoretically underpins much of what goes on in capital structure arbitrage.

If a pension plan fully understood that its leveraged real estate investments were really long term options (warrants) and classified them as such they would have had a far better understanding of the risks embedded in their portfolios. It, therefore follows that, leveraged real estate certainly does not belong in the plain vanilla real estate bucket.

There is, of course, an agency argument for high leverage in REITs and private equity. Briefly stated leverage is required to increase the return on equity that is so necessary to compensate the best talent. That certainly was the argument from 2004-2007. However, many of the most talented real estate professionals crashed and burned with everyone else in 2008. It seems to me that the agency argument is better suited for bull markets than bear markets. All told, then, the evidence suggests that less leverage is preferable to more.

*David Shulman was formerly the Senior REIT Analyst at Lehman Brothers. He is now affiliated with Baruch College, the University of Wisconsin and the UCLA Anderson Forecast. He can be contacted at david.shulman@baruch.cuny.edu.