← →

The Problem with Leverage

June 14th, 2009 · 11 Comments

I’ve always been under the impression that leverage will likely be profitable over the long run — say a period of 10 or 20 years (you can find past posts on leverage here, here and here). Stocks beat bonds and cash handily over the long term, right? But is it really the case? What does market history tell us? To find the answer, we’ll turn to Triumph of the Optimists, in which authors Elroy Dimson, Paul Marsh and Mike Staunton present market returns for stocks, bonds and bills from 1900 to 2000 for 16 different countries including Canada and look at excess returns of stocks over bills. As an investor borrowing at 3% and earning 7% in the stock market, only gets to keep the difference, it makes sense to look at the equity risk premium versus bills. Assuming, investors can borrow at the T-bill rate, the sequence of excess returns available from stocks for every 10-year period starting in 1900 is:

3.5%, 2.4%, 8.8%, -0.5%, 8.4%, 12.9%, 5.2%, 2.5%, 0.0%, 3.5%

Unfortunately, investors are not in the lucky position of the Government of Canada and are unlikely to be able to borrow at the current T-bill rate of 0.25%. The best they can usually do is borrow at Prime, which is currently 2.25% or a 2% premium over the T-bill rate. Let’s deduct a premium of 2% from the equity premium sequence above to find out how successful leverage would have been in past 10-year periods:

1.5%, 0.4%, 6.8%, -2.5%, 6.4%, 10.9%, 3.2%, 0.5%, -2.0%, 1.5%

In other words, in 2 out of 10 10-year periods, leverage would have been unprofitable and barely profitable in 2 more. It is only in 4 out of 10 periods that leverage would have yielded substantial profits.

How about 20-year periods? Here’s the sequence of excess returns from stocks over bills after deducting a 2% premium:

0.9%, 3.6%, 2.1%, 1.9%, 8.6%, 7.0%, 1.9%, -0.8%, -0.2%

Again, the record is clear. Leveraging would have provided outsized returns in some periods and mediocre or quite poor results in others.

It is only over 30-year periods that all excess returns are positive:

2.9%, 1.5%, 3.5%, 4.8%, 6.8%, 4.8%, 0.5%, 0.1%

Leverage is typically advertised as a profitable strategy over the long term. But market history suggests that investors may sometimes have to wait a long time before the strategy turns profitable.

Bookmark:  

Buzz Up Digg Mixx
Print

Related Posts:

Tags: Investing

11 responses so far ↓

  • 1 Cash Canuck // Jun 15, 2009 at 2:36 am

    Isn’t there a lack of data? There are a lot more than 10 ten-year periods between 1900 and 2000 (ninety by my count). With fluctuation between 12.9% and -0.5% it’d be worth having a few more numbers to work with.

    Also, how can one calculate the excess returns for leveraging? Do the authors make any assumptions about the leverage ratio? If you leveraged ten-to-one, for example, your excess returns (and losses) would be astronomically greater than an equivalent non-leveraged portfolio or even a two-to-one ratio… so many questions. I should find that book and read it I guess :P .

  • 2 Phil S // Jun 15, 2009 at 9:24 am

    Which stocks were used as measurement? For myself, I usually only leverage to buy stocks that can “carry themselves” for leveraging like REITs, Limited Partnerships and the now-near defunct Income Trusts. In those cases, you can find many stocks yielding over 10% these days, so in my case at Prime +1 (3.25%), the pre-tax spread is 6.75%! And you’re making that 6.75% spread every month.

    In my case, I only worried about the credit quality of the stock to pay cash to cover the interest. One of my stocks doubled in price so I sold it off and paid back all of the leverage and now the OTHER stocks I bought with leverage are essentially “free” to me now even though many of those ones are under water. Despite being under water, they’re still distributing and that’s all I care about – the free cash every month.

  • 3 Canadian Capitalist // Jun 15, 2009 at 10:19 am

    Cash Canuck: I’m assuming the entire portfolio is leveraged, perhaps with home equity as collateral. That is why, only the difference between stock returns and T-bill returns matters. I also agree that the data is fairly sparse — there are only 10 non-overlapping 10-year periods between 1900 to 2000. Still, there are some surprising results here. For instance, in the 20-year period ending in 2000, leverage would have been unprofitable.

    Phil: My point isn’t that leverage never works. In the return sequence above there are periods when leverage would have provided out-sized returns. Perhaps, this is one of those periods. What surprised me is that leverage doesn’t always result in a profit, even over periods as long as 20 years.

  • 4 m1234 // Jun 15, 2009 at 11:19 am

    Interesting study, would like to read it….I would also like to see a comparison to see how leveraging in real estate has paid off in the last 100 years….everyone says borrow to buy a house…I would like to know how profitable this really is (taking out the mental benefits of owning a home and sticking to just the numbers)…and how it would compare to leveraging stocks….

  • 5 Guy Davis // Jun 15, 2009 at 11:35 am

    Did you account for taxation? Ignoring the tax deductions from the investment loan and preferentially treated Canadian dividend stocks when comparing to interest-bearing investments would introduce error in your analysis.

  • 6 Canadian Capitalist // Jun 15, 2009 at 11:53 am

    @Guy Davis: The results are gross before taxes. Tax effects are not taken into consideration. It is true that interest on the investment loan results in a tax deduction and dividends and capital gains are taxes preferentially. However, it is very hard to do an after-tax analysis because tax rates have been all over the map. Tax on only half the capital gains is (if I recall correctly), a post-2000 phenomenon. And I’ve lost count of the number of times dividend taxes have been tinkered with in just the past few years.

    In any case, the capital gains taxation is a tax on nominal gains, not real returns. If I make an $100K investment and inflation runs at 2% and I have $120K at the end of 10 years, I haven’t made any real gains. But if I sell, I’ll be on hook for capital gains tax on $20K of my “profits”. What I’m trying to get at is analyzing after-tax returns isn’t very simple.

  • 7 Four Pillars // Jun 15, 2009 at 12:20 pm

    1) I’m not sure why this is a surprise. Leverage magnifies returns and obviously not all 10 year equity returns are positive – especially after paying interest. I would think 20 years at a minimum is the “long run” – 30 years would be better.

    2) Saying that using leverage isn’t always profitable is like saying that owning stocks isn’t always better than bonds. However true that may be – it’s the overall return for the duration of the investment that matters – not whether some periods were profitable/unprofitable.

  • 8 Canadian Capitalist // Jun 15, 2009 at 1:06 pm

    @Mike: In two out of nine 20-year periods, leverage would have resulted in a loss is certainly a surprise for me. It is especially surprising that the 1980-2000 period turned out to be negative. It is true that only the return for the total holding period counts but investors are likely to be very discouraged if their returns were negative over 20 years. It would be interesting to see the results when the post-2000 horrible period for stocks is included in the analysis.

  • 9 Fred (ETF2X.com) // Jun 15, 2009 at 4:35 pm

    I have to agree with Cash Canuck. When you are backtesting, and I have done my share of it, it is preferred to use sliding windows. If you want to be extraordinarily thorough, you would run tests starting on January 02 and extending for ten years. Then you would run tests starting on January 03 and extending for ten years. In other words, test for sliding windows of ten year periods.

    Of course this analysis is based on buying-and-holding which is a strategy that I don’t employ.

  • 10 Winner Announcement- New Giveaway and Weekly Blog Review // Jun 19, 2009 at 4:24 am

    [...] Canadian Capitalist looks at the problems with leverage [...]

  • 11 Weekend Reading - July 19, 2009 | Million Dollar Journey // Jun 19, 2009 at 6:31 am

    [...] Canadian Capitalist writes about the problem with leverage. [...]

Leave a Comment