Introduction to the SPY-TLT Universal Investment Strategy (UIS)
This paper discusses the simple but effective method of using adaptive allocations between stock market ETFs and Treasuries to assemble a simple yet smart Investment Strategy. This method has been developed to replace the 100% switching used in normal rotation strategies like the Maximum Yield Rotation and the Global Market Rotation strategies.
The real world is just not a 100% “risk on” or “risk off” world. Most of the time, the best allocation is somewhere in between. The new method employed in this investment strategy can be adapted for nearly all types of rotation strategies and is significantly increasing the return to risk (Sharpe) ratio of such strategies.
The SPY-TLT Universal Investment Strategy is very simple but also very effective. I am sure, such a simple investment strategy will nearly always perform better than any manual asset picking.
The Universal Investment Strategy
Probably the most basic rotation investment strategy, is the switching strategy between the S&P 500 US stock market (SPY) and long duration Treasuries (TLT). The SPY-TLT ETF pair is a very interesting investment strategy, because most of the time these two ETFs profit from an inverse correlation. If there is a real stock market correction, then Treasuries like TLT have always been the assets where money flows in, rewarding holders with nice profits.
Now there are two possibilities to profit from this inverse correlation. The first is a switching strategy, which always switches to the ETF which had the best performance during the previous 3 months. This really simple switching strategy between TLT and SPY gave you a 14.8% return during the last 10 years, with twice the Sharpe ratio (return to risk) ratio of a simple SPY investment.
Another strategy would be to invest 50% of your money in SPY and 50% in TLT and do a monthly rebalancing. This gave you 8.8% return during the last 10 years.
Now let’s see, how these two strategies performed the first 10 month of 2014. For quite some time we now had a sort of a sideways market. Conflicts like Syria or Ukraine made investors switch several times in “risk off” mode which favors safe haven assets like our TLT Treasury, but shortly after they switched back to “risk on” mode favoring our SPY stock market ETF. Such whipsaw market is bad for rotation strategies and as a result our SPY-TLT switching strategy only made 5.6% during the first 10 months with a Sharpe ratio of 0.59.
This time the much better strategy was to hold equal positions of SPY and TLT. This strategy made 16.7% during the first 10 months with an very high Sharpe ratio of 3.12.
If you analyze these two strategies, then you will quickly see which are the advantages and disadvantages of each strategy. The switching strategy did well during market periods with a strong up or down trend. In 2008 this strategy could avoid the stock market crash by switching nearly 100% to Treasuries. During sideways market periods and especially during whipsaw market periods, the switching strategy switched a lot of times too late, which resulted in an unwanted sell low and buy high strategy.
Variations of the Universal Investment Strategy
The idea for this Universal Investment Strategy was to develop a strategy which has an adaptive allocation between 0% and 100% for each ETF depending of the market situation. I am using such a strategy for quite some time with excellent results. Since 2014 the strategy invests always approximately equal parts of TLT and SPY.
The way to calculate the optimum composition is done by calculating which composition had the maximum Sharpe ratio during an optimized look back period (normally 50-80 days). During normal market periods, the maximum Sharpe ratio is not at a 100% SPY or at a 100% TLT allocation, but somewhere in between. To calculate this maximum Sharpe ratio, I loop through all possible compositions from 0%SPY-100%TLT to 100%SPY-0%TLT and calculate the resulting Sharpe ratio for the look back period.
As a result I get a curve like this (result of July 21, 2014):
The interesting finding is, that the best combination of SPY-TLT gets with 5.27 a considerably higher Sharpe ratio then SPY (3.38) or TLT (2.01) alone. This is because the inverse correlation of the two ETFs reduces volatility or risk a lot.
On July 21, 2014 when I made this calculation, this would mean that the optimum composition for the portfolio would be 60% SPY and 40% TLT. This composition is in this case also the one with the lowest volatility.
Year to date the volatility of such a strategy is below 5%, which is less than half the volatility of SPY or 3x to 4x less compared to other markets (Europe, Emerging markets).
For my UIS strategy I tweaked the Sharpe formula a little bit. Normally the Sharpe ratio is calculated by Sharpe = rd/sd with rd=mean daily return and sd= standard deviation of daily returns. I don’t use the risk free rate, as I only use the Sharpe ratio to do a ranking. My algorithm uses the modified Sharpe formula Sharpe = rd/(sd^f) with f=volatility factor. The f factor allows me to change the importance of volatility.
If f=0, then sd^0=1 and the ranking algorithm will choose the composition with the highest performance without considering volatility.
If f=1, then I have the normal Sharpe formula.
If f>1, then I rather want to find SPY-TLT combinations with a low volatility. With high f values, the algorithm becomes a “minimum variance” or “minimum volatility” algorithm.
To get good results, the f factor should normally be higher than 1. This way you do not need to rebalance too much. In a whipsaw market, rebalancing also has the negative effect of selling low and buying high on small intermediate market corrections. This is why a system which considers only performance will not do well.
The good f factor for a system can be found by “walk forward” optimization iterations of your backtests. Normally a good value for f is about 2, but the factor changes slightly, adapting to the current market conditions.
Comparison of different SPY-TLT investment strategies (5 year)
|Strategy||5 year CAGR||Sharpe ratio|
|1||SPY-TLT UIS adaptive allocation||18% annual return||1.66|
|2||50% SPY- 50% TLT rebalanced monthly||12.5% annual return||1.31|
|3||SPY – TLT monthly rotation strategy||19% annual return||1.19|
It is interesting to see, that the return of such an investment strategy will be higher if market volatility is higher. So, for the low volatility period from 2000-2007 (SPY volatility of 13), the annual return was about 9%, but since then, we had an average SPY volatility of 23, which results in a much higher return for the strategy. So this means, that with such a strategy we are happy if we have a really nice 25% market correction followed by a recovery, because this means, that we can profit from the market corrections by over-weighting TLT and later on, we profit again from the recovery by over-weighting SPY.
So, in summary, a variable allocation between stock market ETFs and Treasuries has by far the best risk/return (Sharpe) ratio for this investment strategy. Fixed ratios do only perform well, when market conditions do not change. Be skeptical if somebody will sell you a strategy with a fixed allocation which is only backtested for the last few bull market years. Normally these strategies will fail when the next 2008 like bear market begins. The older methodology SPY-TLT rotation has a slightly higher return, but a much higher volatility or risk. Return alone should never be used to judge a strategy. In fact it is very simple to boost the return by using leveraged ETF replacements for SPY and TLT. But this would in no way make the strategy superior, because then also the volatility or risk will be higher by the leverage factor.
Here is a plot of the results if you use short positions of the -3x leveraged SPXS-TMV instead of SPY-TLT. The result is incredible. 2051% return for the 5.5 year period. This is a 74% annual return with a Sharpe ratio of 2.6. In the below chart, you see also a green chart which shows you the monthly SPXS allocation. The TMV allocation is always the 100%- SPXS allocation. It is interesting to see, that even if we had a bull market during this period, in average the strategy is about 50% invested in Treasuries.
Other uses for the Universal Investment Strategy
The adaptive SPY-TLT allocation strategy is also very interesting, because you can use it for different other investment strategies. In general the UIS strategy gives you the very important information, on how to position your investments between “risk on” stock market investments and “risk off” Treasuries. Instead of using this to invest directly in the SPY and TLT ETFs, you can us this allocation for a lot of other ETFs.
Here are some Investment strategies based on the calculated UIS allocation:
|Strategy||5 year CAGR||Sharpe ratio|
|1||Basic ETF strategy using SPY and TLT ETFs||18%||1.66|
|2||Futures strategy using S&P 500 ES and UB or ZB Treasury Futures||19%||2.25|
|3||Short positions of -3x leveraged SPXS and -3x leveraged TMV||74%||2.6|
|4||Option strategies. Selling -2 SD SPY put options and -1SD TLT put options||About 75% on margin requirement||??|
Explanation of the strategies.
- This is the basic strategy which invests in the SPY and the TLT ETFs. Both are very liquid ETFs with tight spreads and low commission costs. If you have a margin account, then there is no problem to invest with a leverage of about 2x to 3x because of the low volatility of the strategy. This way you can easily have returns of up to 50% with your money.
- I like the futures strategy, because futures are the most efficient instruments to trade the S&P 500 and Treasuries. You only need about 2-4% margin to buy these futures, so in theory you could have a leverage up to 20x. Commissions are also very low. A big advantage compared to ETFs is that Futures trade nearly 24h a day. This allows you to exit your positions at a good price in case of a sudden crisis before the ETF market opens. Another advantage is, that Treasury futures don’t pay dividends. Dividend is included in the Future price and results in a capital gain when you roll the futures. For example in Switzerland this is a big advantage, because capital gains are tax free and dividends are taxed like a normal income.
In the States futures are submitted to the 60%-40% tax rule, which means that 60% of a capital gain or loss may be treated as a long-term capital gain or loss and 40% may be treated as a short-term capital gain or loss, even if the position was held for less than a year.
A disadvantage with futures is, that they usually expire after one to three months. Therefore, you have to roll (sell the old and buy the new future) them every 3 month. Another disadvantage is that one future is normally about 100’000 US$ or more. So, if you only have a small account, then futures are too big for you.
- If you can short ETFs, then I would invest in the -3x leveraged equivalents of SPY and TLT. This is what I do since about 2 years. If you short these ETFs, then you are in fact long with a 3x leverage. The nice thing is, that these -3x leveraged ETFs have very big rebalancing losses and high commissions. If you short these ETFs, then you profit from these losses. This adds about 1% of free profit per month to your strategy. Both TMV and SPXS have quite low borrowing costs of normally below 3% per year. This type of strategy is also known under the name “Hedged Convexity Capture”. Most of these strategies tell you to invest in fixed ratios of -3x leveraged stock and Treasury ETFs. But attention! This worked well the last years of the bull market, but as you can backtest with the 50%SPY-50%TLT strategy, such a “fixed ratio” strategy would have resulted in a loss of near to 100% during the 2008 subprime crash.
- Another interesting strategy is selling about 1-3 standard deviations OTM (Out of the Money) put options with about a 50 day expiry on SPY and TLT. In addition to the about 82% probability of a winning month for the base SPY-TLT strategy, you invest in options with a more than 90% probability of profit. This strategy is very interesting, because even if you once sell for example the SPY put option with a loss, then you can be nearly 100% sure that the other TLT put option will return the maximum profit.
There are two advantages over normal option strategies. Normally if you calculate a probability of profit for such a “naked” short put trade, the probability is trend neutral, only depending of implied volatility. This would mean that you have a 50% probability for an ETF to go up or down. However if you add our SPY-TLT UIS basic strategy to it, then you increase the probability of profit, because you will invest with a high probability trend. The other advantage is to execute always pair trades with an inverse correlation. Together with stop loss orders, this reduces the remaining risk a lot.
This strategy has a low volatility because both options have about the same delta and therefore inverse moves. The thing which you will see, is the slow premium (i.e., theta) decay adding profit to your account. I am still testing this strategy and while I could backtest the monthly returns I was not able until now to calculate volatility or Sharpe ratios. However the volatility seems well below 5%.
Another possible option strategy is to sell ATM (At the Money) put options with an expiration of about 4-6 months. If you hold them only for about 2-3 month, then delta will not change much and such a position behaves much like a long position of SPY or TLT within about +-5% price changes of the underlying ETF. The difference to a direct investment in SPY-TLT is, that here you get the additional premium which will boost your profits. Drawback of these option strategies is the low volume of traded TLT options which sometimes results in large bid-ask spreads.