Thursday, May 21, 2009

Are Leveraged ETFs Leaky?

The basic idea behind leveraged ETFs seems appealing. These ETFs seek to go up or down double or triple the amount of a given index. Since stocks tend to go up over time, you’d think that doubling or tripling this gain would be a good idea. Unfortunately, volatility punishes leveraged ETFs and some of these ETFs seem to have unexplained leaks.

Many investors believe that if an index rises 10% one year, then a double-leveraged ETF will rise 20%, and a triple-leveraged ETF will rise 30%. If you think this is true, you need to slow down and understand where this reasoning goes wrong. These ETFs get rebalanced daily and this changes everything.

Example

Suppose that a stock index alternates between rising 2% one day and dropping 1.9% the next day for 250 trading days in a year. Superficially, you might think that you are gaining 0.1% every two days repeated 125 times giving a 12.5% gain over the year. Sadly, it doesn’t work this way. Over 2 days, a $1000 investment would be affected as follows:

$1000 x 1.02 x 0.981 = $1000.62

Note that the gain is only 0.062% rather than 0.1%. For the double- and triple-leveraged ETFs, we have

2X: $1000 x 1.04 x 0.962 = $1000.48
3X: $1000 x 1.06 x 0.943 = $999.58

Multiplying this out 125 times for the whole year gives the following results:

1X: +8%
2X: +6%
3X: -5%

The leveraged ETFs magnify volatility, and this compounded volatility punishes returns.

All this analysis has been based on the “bull” ETFs. There are also “bear” ETFs that seek to get the opposite return of the index. So, if the index goes down 1% one day, a double-leveraged bear ETF would go up about 2%.

A natural question is “would a triple-bear ETF go up 5% in the example above?” Sadly, the answer is no. The volatility penalty makes both the bull and bear ETFs go down. Here are the bear returns for this example:

2X bear: -36%
3X bear: -55%

These returns have excluded all trading fees, MERs, and any other expenses.

An Experiment

Is this volatility problem the only drain on leveraged ETF returns? I decided to try an experiment where a pot of money is invested 50/50 between the bull and bear versions of the same ETF. With daily rebalancing, this should eliminate the volatility problem. The gain in one ETF should cancel the loss in the other each day leaving just expenses.

Of course, it makes no sense for an investor to actually do this. The goal of this experiment is to eliminate volatility losses and see what happens. My analysis ignores the trading fees that would result from the daily rebalancing because I want to examine the nature of these ETFs.

The first ETFs that I tried this on were the Direxion US Large Cap triple bull and bear. The following chart shows what would have happened to $10,000 split between these ETFs since their inception.



If we ignore the initial big drop, these ETFs seem to be leaking money at a consistent rate of about 9% per year. This far exceeds the advertised expenses of about 1%. The bull ETF also paid out a couple of small dividends, but this falls far short of explaining the leak. Perhaps, the trend in the chart is only short-term and will change. In the end, I have no explanation for this drop.

The results for Canadian Horizon BetaPro TSX 60 double-exposure bull and bear ETFs were stranger:



It’s hard to see any particular trend in this data. The overall drop of about 3.5% in 20 months exceeds the advertised yearly MER of 1.15%. The inconsistency of results seems to indicate that the ETFs do not track the index exactly as intended, but it’s hard to tell.

Between the volatility penalty and the unexplained leaking of money when volatility is factored out, these ETFs won’t be part of my portfolio.

8 comments:

  1. Michael, excellent post ... but you are not done. Your task is now to figure out why your two experimental portfolios performed as they did. Something important for investors is hidden within.

    ReplyDelete
  2. Canadian Investor: I was hoping that my readers might have a theory or two. I'm at a loss to explain this apparent leak.

    ReplyDelete
  3. Really well-researched and calculated post. This belongs in a journal of finance.

    The Thicken My Wallet post you linked was helpful in explaining the rebalancing mechanism that throws a wrench into the works. After reading the first part of his post, I also thought the ETF's systematic risks are not inconsequential, either. Their use of derivatives could conceivably cause some kind of financial shock to the ETF's stability.

    I agree with your avoidance of these funds.

    ReplyDelete
  4. Well done. Thank you for your thought provoking research. It's always informative and a pleasure to read your blog.

    I hope you keep up the good work.

    ReplyDelete
  5. Okay, this begs the question: short both the bull and bear in equal amounts? Does this capture the fees and drag in your favour?

    Great post, absolutely fantastic post.

    ReplyDelete
  6. Most investment advisors that I've spoken with (and many that are featured on BNN) advise people to stay away from these ETFs as well. The ones that do recommend them basically say that they should be used for, at most, a day trade. Since the vast majority of investors cannot predict the direction of the market on a day to day basis I would consider investing in these ETFs for a day to be gambling.

    ReplyDelete
  7. the only people who recommend these crappy ETFs are the people who manage them and put the MER in their pockets.

    ReplyDelete
  8. Preet: I thought about shorting the bull and bear, but I'm not sure that the gains would be enough to offset the shorting costs (borrowing shares and maintaining margin) if you maintain a balance in the dollar amount of each ETF. You could try not rebalancing, but this adds risk to an otherwise mostly riskless arbitrage.

    Anonymous: I agree that short-term investing is essentially gambling.

    ReplyDelete