3x ETFs: Using Leverage For Higher Returns

Much has been written about why 3x long ETFs are not for long term investments. One of the reasons given is that both the 3x long ETF and the 3x short ETF have losses since their inception.

Though I agree with the conclusion, I disagree with the evidence given. The performance of the ETFs was during an extremely volatile sideways and bear market and not during a bull market. Because I do not consider the evidence given conclusive, I thought that I would use more data to get to that conclusion and show how 3x leveraged ETFs could be used by intermediate-term investors. Though I consider it possible for investors to use 3x leveraged ETFs, this is by no means a recommendation for anyone to use them.

First, there is not enough historical data from the 3x ETFs to include a bull market, so I need to use index data and triple the daily performance. However, before I use an index I have to make sure the ETFs’ price performance actually tracks triple the daily performance of their indices. Here is the Direxion Large Cap Bull 3x Shares compared with 3 times the daily performance of its index, the Russell 1000 Index.

As you can see, the ETF tracks its index almost perfectly. The other 3x ETFs also have similarly high correlation. I cannot prove that the correlation will hold for a longer period as there is not much data, but the correlation is high enough to make me confident that it will hold over the longer term and that I can use a hypothetical 3x leveraged index as a good proxy for how the ETF would have performed without management fees.

Here is a chart with the Russell 1000, the hypothetical 2x Russell 1000, and the hypothetical 3x Russell 1000.

As you can see, during a bear market, 3x long ETFs will have huge losses, while during bull markets they will have strong gains.

The chart above is probably convincing enough to deter anyone from buying and holding the 3x long ETFs, but one can argue that the time period includes more bear market than normal. Therefore, I created a chart of the S&P 500 and hypothetical 2x and 3x leveraged S&P 500.

This time the chart is not as convincing that buying and holding 3x long ETFs is a bad idea. It does not convince me. Sure, it has several enormous losses, but in the next bull market it always makes up for its losses. If it did not convince you either, than the next one will. This chart is of the Dow Jones Industrial Average (DJIA), 2x DJIA (hypothetical), and 3x DJIA(hypothetical) since 1986.

After the DJIA peaked on September 3, 1929 and subsequently crashed, it took nearly 58 years for the 3x DJIA to gain all the losses back. Just sit and let that sink in. If you had started saving for retirement and bought a 3x long DJIA ETF at the top in 1929 when you were 20, you would be nearly 78 before it would have gained back its losses. Then just as you started making a little money and thinking you could finally retire, it subsequently crashes over 60% in one day. Welcome to stocks for the super long run.

Use of leveraged ETFs

Since it does not seem like a good idea to buy 3x ETFs and hope to retire, I thought I would explore some moving average crossover strategies and see how well they worked with leveraged indices. First, I back tested a 20EMA 50EMA crossover strategy and a 50EMA 200EMA crossover strategy on the DJIA. Then I back tested investing only when either the 20EMA is above 50EMA or the 50EMA is above the 200EMA. If a signal was given at the close of one day than the signal would be acted on (bought or sold) at the close of the next day.

Since a moving average crossover strategy is not a good strategy unless it works on more than one index I back tested the same three strategies on the S&P 500, Nasdaq Composite, and Russell 2000.

Now I need to rank the three strategies to find out which one is best. For each index the crossover system is given a three if it outperformed the other systems, a two if it outperformed only one, and a one if it underperformed the other two systems. The scores are than added up.

1. 20EMA > 50EMA: 2+1+3+2=8
2. 50EMA > 200EMA: 1+2.5+1+1+5.5
3. 20EMA > 50EMA OR 50EMA > 200EMA: 3+2.5+2+3=10.5

Clearly, strategy three wins. Next, I back test investing in leveraged indices when the strategy applied to a non-leveraged index signals a buy.

Without leverage, this strategy has similar returns and lower peak to trough losses than the index. With 2x leveraged indices the returns are substantially higher and the peak to trough losses are slightly greater than the index. And for 3x leveraged indices, the return is even higher while the peak to trough losses are substantially higher than the index. Though this strategy is crude, I used because it is simple and easy to understand. It is likely that if criteria such as stock market valuation, interest rate trends, and bond risk spreads are added, higher returns or lowered risk could be achieved.

I publish the signals from a model in my free monthly newsletter at PlumbobInvestments.com that can be used to signal when to use leverage. It signals sell, hold / moderate buy, or strong buy. The strong buy signal is a highly reliable signal and I recommend the prudent use of leverage when it gives a strong buy signal.

In the end, it all comes down to the timing of the purchase of a 3x long ETF. If you buy before the beginning of a secular bull market, than even though the volatility of your investment may be higher than the index, your returns would be much higher than the index. If you buy before a secular bear market, you may end up waiting many years before your investment gains its value back. In fact, if I reduce the peak to trough loss of the 3x leveraged DJIA during the Great Depression to the peak to trough loss of the simple strategy I outlined (-96% Still a larger peak to trough loss than the DJIA), the returns of the 3x leveraged DJIA would have by far outperformed the non-leveraged DJIA.

I still would not recommend using 3x leveraged ETFs for intermediate-term investments unless you have a good rules based model, the ability to control your emotions, and money you can risk losing. You need a model to limit the losses of the leveraged ETF. However, the model is only useful if you follow it. Your emotion might cause you to sell out and stop using your strategy after it lost a lot. If you stop using 3x leveraged ETFs after losing a lot, there is almost no way to gain it back unless you continue to use leverage. Lastly, you should only use money you can afford to lose because your model might have a bad losing streak and you might sell out or stop using it.

Shorting Leveraged Long and Short ETFs

The fact that some writers have pointed out that both the 3x long ETFs and the 3x short ETFs have lost money since inceptions has led some to suggest shorting both the long and the short ETFs. Besides the fact that the shares may be hard or impossible to borrow, I do not think this is a good idea. Doing so recently may have earned you money, but it may not have always done so in the past and may not do so in the future. To understand my point, look at the charts below. The reason these are not rebalanced daily is because the ETFs track their indices very closely. If one track 300% of the index and the other 300% of the inverse of the index, than rebalancing daily would have the effect of cancelling out any return. Again, these are based on hypothetical 3x leveraged indices and not actual ETFs.

It certainly does not look like you would want to short both the long and short leveraged ETFs. The reason shorting both worked recently is because of the recent extreme volatility. When the stock market goes up 5% one day and then declines 5% the next day, the stock market actually loses 0.25%. However, the 3x long ETF will go up 15% and then decline 15% resulting in a 2.25% loss while the 3x short ETF will also end with a 2.25% loss. This explains why both the long and short ETFs have lost since inception. Next, consider a three day period when the stock market rises 1% each day. At the end of the period the 3x long ETF will have gained 3.0301% and the 3x short ETF will have lost 2.9701%. If the stock market declined 1% three days in a row the same would happen except in reverse with the short ETF rising 3.0301% and the long ETF declining 2.9701%. So if the stock market is volatile but moves in one direction over the period you shorted both ETFs than one of the ETFs may have risen more than the other declined and you would have a larger loss in one than a gain in the other. The only reason to short both ETFs is if you expect volatility to be very high and you don’t expect the stock market to rise or decline much during the period you plan to hold the position. Historically there have been very few times when conditions were just right to short both the long and short ETFs.

4x Leverage?

Next up after3x leverage; 4x leverage. Just in case someone comes up with the idea to create 4x leveraged ETFs I thought I’d make a chart with a 4x daily index of the DJIA.

Suddenly when compared to the 4x leveraged index, the 3x leveraged index doesn’t look quite so horrible anymore. If you bought a 3x long ETF at the top in 1929 it would have taken you at least a mere 58 years before you were back to even, but with a 4x leverage you would never have recovered your losses.

Negative Value?

Speaking of super leveraged indices and ETFs, what if there was a 5x ETF and the stock market crashed 21% in one day? The ETF should drop by 105%. 5 x 21%= 105%. I wonder what would happen. Would the value turn negative? Or with the current 3x Long ETFs, what if the stock market crashed 35%? 3 x 35% = 105%. Would it be quoted as trading at -$5 or something of the like?

Disclosure: No positions

Comments

[49er:]

Thanks for the analysis.

Apr 17 08:31 AM

[hksche2000:]

Great work, thanks a mio!
FAS has done very well short term lately for me, but you can't ride it max and still sleep well. Vegas revisited.

Apr 17 01:22 PM

[rogerk2:]

Regarding a series of 35% crashes (3 x 35% = 105%) -- the 3X ETF wouldn't ever reach zero. Say it's $10. A 35% loss would rebalance the price to $6.50 the next day. Another 35% loss would get rebalanced to $4.23. Another 35% loss would take it to $2.75, and so forth. Since today's price is a percentage increase or decrease based on yesterday's price, it could get close to zero but would never actually reach it.

FAS and FAZ have had lots of 30% and 40% daily moves, which is why they are so popular with daytraders and often have the highest daily volume on the exchanges. Play them with extreme caution. The professional traders have access to market-moving information before you do, they have enough money to manipulate any stock they want to, and they can trade before and after your markets open/close. If you play against them, you lose.

Apr 18 01:14 AM

[Northstar:]

The point on 3X ETFs are not buy and hold anyway. Cabs are expensive but is well worth it when you need one and its fare is not comparable with your own car.

I would sell my ETF immediately once I smell a down trend.
Thank you for the analysis.
S

Apr 18 02:12 AM

[Geoffrey Ching:]

rogerk2

Yes thats true for 35% moves of the ETF, but what if the index moves 35% in one day? 35% x 3= 105%.What will the ETF do then?

Apr 23 10:34 AM