Statistical Arbitrage with Synthetic Data

In my last post I mapped out how one could test the reliability of a single stock strategy (for the S&P 500 Index) using synthetic data generated by the new algorithm I developed.

Developing Trading Strategies with Synthetic Data

As this piece of research follows a similar path, I won’t repeat all those details here. The key point addressed in this post is that not only are we able to generate consistent open/high/low/close prices for individual stocks, we can do so in a way that preserves the correlations between related securities. In other words, the algorithm not only replicates the time series properties of individual stocks, but also the cross-sectional relationships between them. This has important applications for the development of portfolio strategies and portfolio risk management.

KO-PEP Pair

To illustrate this I will use synthetic daily data to develop a pairs trading strategy for the KO-PEP pair.

The two price series are highly correlated, which potentially makes them a suitable candidate for a pairs trading strategy.

There are numerous ways to trade a pairs spread such as dollar neutral or beta neutral, but in this example I am simply going to look at trading the price difference. This is not a true market neutral approach, nor is the price difference reliably stationary. However, it will serve the purpose of illustrating the methodology.

Historical price differences between KO and PEP

Obviously it is crucial that the synthetic series we create behave in a way that replicates the relationship between the two stocks, so that we can use it for strategy development and testing. Ideally we would like to see high correlations between the synthetic and original price series as well as between the pairs of synthetic price data.

We begin by using the algorithm to generate 100 synthetic daily price series for KO and PEP and examine their properties.

Correlations

As we saw previously, the algorithm is able to generate synthetic data with correlations to the real price series ranging from below zero to close to 1.0:

Distribution of correlations between synthetic and real price series for KO and PEP

The crucial point, however, is that the algorithm has been designed to also preserve the cross-sectional correlation between the pairs of synthetic KO-PEP data, just as in the real data series:

Distribution of correlations between synthetic KO and PEP price series

Some examples of highly correlated pairs of synthetic data are shown in the plots below:

In addition to correlation, we might also want to consider the price differences between the pairs of synthetic series, since the strategy will be trading that price difference, in the simple approach adopted here. We could, for example, select synthetic pairs for which the divergence in the price difference does not become too large, on the assumption that the series difference is stationary. While that approach might well be reasonable in other situations, here an assumption of stationarity would be perhaps closer to wishful thinking than reality. Instead we can use of selection of synthetic pairs with high levels of cross-correlation, as we all high levels of correlation with the real price data. We can also select for high correlation between the price differences for the real and synthetic price series.

Strategy Development & WFO Testing

Once again we follow the procedure for strategy development outline in the previous post, except that, in addition to a selection of synthetic price difference series we also include 14-day correlations between the pairs. We use synthetic daily synthetic data from 1999 to 2012 to build the strategy and use the data from 2013 onwards for testing/validation. Eventually, after 50 generations we arrive at the result shown in the figure below:

As before, the equity curve for the individual synthetic pairs are shown towards the bottom of the chart, while the aggregate equity curve, which is a composition of the results for all none synthetic pairs is shown above in green. Clearly the results appear encouraging.

As a final step we apply the WFO analysis procedure described in the previous post to test the performance of the strategy on the real data series, using a variable number in-sample and out-of-sample periods of differing size. The results of the WFO cluster test are as follows:

The results are no so unequivocal as for the strategy developed for the S&P 500 index, but would nonethless be regarded as acceptable, since the strategy passes the great majority of the tests (in addition to the tests on synthetic pairs data).

The final results appear as follows:

Conclusion

We have demonstrated how the algorithm can be used to generate synthetic price series the preserve not only the important time series properties, but also the cross-sectional properties between series for correlated securities. This important feature has applications in the development of statistical arbitrage strategies, portfolio construction methodology and in portfolio risk management.

Machine Learning Based Statistical Arbitrage

Previous Posts

I have written extensively about statistical arbitrage strategies in previous posts, for example:

Applying Machine Learning in Statistical Arbitrage

In this series of posts I want to focus on applications of machine learning in stat arb and pairs trading, including genetic algorithms, deep neural networks and reinforcement learning.

Pair Selection

Let’s begin with the subject of pairs selection, to set the scene. The way this is typically handled is by looking at historical correlations and cointegration in a large universe of pairs. But there are serious issues with this approach, as described in this post:

Instead I use a metric that I call the correlation signal, which I find to be a more reliable indicator of co-movement in the underlying asset processes. I wont delve into the details here, but you can get the gist from the following:

The search algorithm considers pairs in the S&P 500 membership and ranks them in descending order of correlation information. Pairs with the highest values (typically of the order of 100, or greater) tend to be variants of the same underlying stock, such as GOOG vs GOOGL, which is an indication that the metric “works” (albeit that such pairs offer few opportunities at low frequency). The pair we are considering here has a correlation signal value of around 14, which is also very high indeed.

Trading Strategy Development

We begin by collecting five years of returns series for the two stocks:

The first approach we’ll consider is the unadjusted spread, being the difference in returns between the two series, from which we crate a normalized spread “price”, as follows.

This methodology is frowned upon as the resultant spread is unlikely to be stationary, as you can see for this example in the above chart. But it does have one major advantage in terms of implementation: the same dollar value is invested in both long and short legs of the spread, making it the most efficient approach in terms of margin utilization and capital cost – other approaches entail incurring an imbalance in the dollar value of the two legs.

But back to nonstationarity. The problem is that our spread price series looks like any other asset price process – it trends over long periods and tends to wander arbitrarily far from its starting point. This is NOT the outcome that most statistical arbitrageurs are looking to achieve. On the contrary, what they want to see is a stationary process that will tend to revert to its mean value whenever it moves too far in one direction.

Still, this doesn’t necessarily determine that this approach is without merit. Indeed, it is a very typical trading strategy amongst futures traders for example, who are often looking for just such behavior in their trend-following strategies. Their argument would be that futures spreads (which are often constructed like this) exhibit clearer, longer lasting and more durable trends than in the underlying futures contracts, with lower volatility and market risk, due to the offsetting positions in the two legs. The argument has merit, no doubt. That said, spreads of this kind can nonetheless be extremely volatile.

So how do we trade such a spread? One idea is to add machine learning into the mix and build trading systems that will seek to capitalize on long term trends. We can do that in several ways, one of which is to apply genetic programming techniques to generate potential strategies that we can backtest and evaluate. For more detail on the methodology, see:

I built an entire hedge fund using this approach in the early 2000’s (when machine learning was entirely unknown to the general investing public). These days there are some excellent software applications for generating trading systems and I particularly like Mike Bryant’s Adaptrade Builder, which was used to create the strategies shown below:

Builder has no difficulty finding strategies that produce a smooth equity curve, with decent returns, low drawdowns and acceptable Sharpe Ratios and Profit Factors – at least in backtest! Of course, there is a way to go here in terms of evaluating such strategies and proving their robustness. But it’s an excellent starting point for further R&D.

But let’s move on to consider the “standard model” for pairs trading. The way this works is that we consider a linear model of the form

Y(t) = beta * X(t) + e(t)

Where Y(t) is the returns series for stock 1, X(t) is the returns series in stock 2, e(t) is a stationary random error process and beta (is this model) is a constant that expresses the linear relationship between the two asset processes. The idea is that we can form a spread process that is stationary:

Y(t) – beta * X(t) = e(t)

In this case we estimate beta by linear regression to be 0.93. The residual spread process has a mean very close to zero, and the spread price process remains within a range, which means that we can buy it when it gets too low, or sell it when it becomes too high, in the expectation that it will revert to the mean:

In this approach, “buying the spread” means purchasing shares to the value of, say, $1M in stock 1, and selling beta * $1M of stock 2 (around $930,000). While there is a net dollar imbalance in the dollar value of the two legs, the margin impact tends to be very small indeed, while the overall portfolio is much more stable, as we have seen.

The classical procedure is to buy the spread when the spread return falls 2 standard deviations below zero, and sell the spread when it exceeds 2 standard deviations to the upside. But that leaves a lot of unanswered questions, such as:

  • After you buy the spread, when should you sell it?
  • Should you use a profit target?
  • Where should you set a stop-loss?
  • Do you increase your position when you get repeated signals to go long (or short)?
  • Should you use a single, or multiple entry/exit levels?

And so on – there are a lot of strategy components to consider. Once again, we’ll let genetic programming do the heavy lifting for us:

What’s interesting here is that the strategy selected by the Builder application makes use of the Bollinger Band indicator, one of the most common tools used for trading spreads, especially when stationary (although note that it prefers to use the Opening price, rather than the usual close price):

Ok so far, but in fact I cheated! I used the entire data series to estimate the beta coefficient, which is effectively feeding forward-information into our model. In reality, the data comes at us one day at a time and we are required to re-estimate the beta every day.

Let’s approximate the real-life situation by re-estimating beta, one day at a time. I am using an expanding window to do this (i.e. using the entire data series up to each day t), but is also common to use a fixed window size to give a “rolling” estimate of beta in which the latest data plays a more prominent part in the estimation. The process now looks like this:

Here we use OLS to produce a revised estimate of beta on each trading day. So our model now becomes:

Y(t) = beta(t) * X(t) + e(t)

i.e. beta is now time-varying, as can be seen from the chart above.

The synthetic spread price appears to be stationary (we can test this), although perhaps not to the same degree as in the previous example, where we used the entire data series to estimate a single, constant beta. So we might anticipate that out ML algorithm would experience greater difficulty producing attractive trading models. But, not a bit of it – it turns out that we are able to produce systems that are just as high performing as before:

In fact this strategy has higher returns, Sharpe Ratio, Sortino Ratio and lower drawdown than many of the earlier models.

Conclusion

The purpose of this post was to show how we can combine the standard approach to statistical arbitrage, which is based on classical econometric theory, with modern machine learning algorithms, such as genetic programming. This frees us to consider a very much wider range of possible trade entry and exit strategies, beyond the rather simplistic approach adopted when pairs trading was first developed. We can deploy multiple trade entry levels and stop loss levels to manage risk, dynamically size the trade according to current market conditions and give emphasis to alternative performance characteristics such as maximum drawdown, or Sharpe or Sortino ratio, in addition to strategy profitability.

The programatic nature of the strategies developed in the way also make them very amenable to optimization, Monte Carlo simulation and stress testing.

This is but one way of adding machine learning methodologies to the mix. In a series of follow-up posts I will be looking at the role that other machine learning techniques – such as deep learning and reinforcement learning – can play in improving the performance characteristics of the classical statistical arbitrage strategy.

Is Internal Bar Strength A Random Walk? The Case of Exxon-Mobil

For those who prefer a little more rigor in their quantitative research, I can offer more a somewhat more substantive statistical argument in favor of the IBS indicator discussed in my previous post.

Specifically, we can show quite convincingly that the IBS process is stationary, a highly desirable property much sought-after in, for example, the construction of statistical arbitrage strategies.  Of course, by construction, the IBS is constrained to lie between the values of 0 and 1, so non-stationarity in the mean is highly unlikely.  But, conceivably, there could be some time dependency in the process or in its variance, for instance.  Then there is the further question as to whether the IBS indicator is mean-reverting, which would indicate that the underlying price process likewise has a tendency to mean revert.

Let’s take the IBS series for Exxon-Mobil (XOM) as an example to work with. I have computed the series from the beginning of 1990, and the first 100 values are shown in the plot below.


XOMIBS

 

 

Autocorrelation and Unit Root Tests

There appears to be little patterning in the process autocorrelations, and this is confirmed by formal statistical tests which fail to reject the null hypothesis that the first 20 autocorrelations are not, collectively, statistically significant.

XOMIBS Autocorrelations

 

XOMIBS acf test

 

SSALGOTRADING AD

Next we test for the presence of a unit root in the IBS process (highly unlikely, given its construction) and indeed, unsurprisingly, the null hypothesis of a unit root is roundly rejected by the Dickey-Fuller and Phillips-Perron tests.

 

XOMIBS Unit Root

 

Variance Ratio Tests

We next conduct a formal test to determine whether the IBS series follows a random walk.

The variance ratio test assesses the null hypothesis that a univariate time series y is a random walk. The null model is

y(t) = c + y(t–1) + e(t),

where c is a drift constant (assumed zero for the IBS series) and e(t) are uncorrelated innovations with zero mean.

  • When IID is false, the alternative is that the e(t) are correlated.
  • When IID is true, the alternative is that the e(t) are either dependent or not identically distributed (for example, heteroscedastic).

 

We test whether the XOM IBS series is a random walk using various step sizes and perform the test with and without the assumption that the innovations are independent and identically distributed.

Switching to Matlab, we proceed as follows:

q = [2 4 8 2 4 8];
flag = logical([1 1 1 0 0 0]);
[h,pValue,stat,cValue,ratio] = vratiotest(XOMIBS,’period’,q,’IID’,flag)

Here h is a vector of Boolean decisions for the tests, with length equal to the number of tests. Values of h equal to 1 indicate rejection of the random-walk null in favor of the alternative. Values of h equal to 0 indicate a failure to reject the random-walk null.

The variable ratio is a vector of variance ratios, with length equal to the number of tests. Each ratio is the ratio of:

  • The variance of the q-fold overlapping return horizon
  • q times the variance of the return series

For a random walk, these ratios are asymptotically equal to one. For a mean-reverting series, the ratios are less than one. For a mean-averting series, the ratios are greater than one.

For the XOM IBS process we obtain the following results:

h =  1   1   1   1   1   1
pValue = 1.0e-51 * [0.0000 0.0000 0.0000 0.0000 0.0000 0.1027]
stat = -27.9267 -21.7401 -15.9374 -25.1412 -20.2611 -15.2808
cValue = 1.9600 1.9600 1.9600 1.9600 1.9600 1.9600
ratio = 0.4787 0.2405 0.1191 0.4787 0.2405 0.1191

The random walk hypothesis is convincingly rejected for both IID and non-IID error terms.  The very low ratio values  indicate that the IBS process is strongly mean reverting.

 

Conclusion

While standard statistical tests fail to find evidence of any non-stationarity in the Internal Bar Strength signal for Exxon-Mobil, the hypothesis that the series follows a random walk (with zero drift) is roundly rejected by variance ratio tests.  These tests also confirm that the IBS series is strongly mean reverting, as we previously discovered empirically.

This represents an ideal scenario for trading purposes: a signal with the highly desirable properties that is both stationary and mean reverting.  In the case of Exxon-Mobil, there appears to be clear evidence from both statistical tests and empirical trading strategies using the Internal Bar Strength indicator that the tendency of the price series to mean-revert is economically as well as statistically significant.

The Internal Bar Strength Indicator

Internal Bar Strength (IBS) is an idea that has been around for some time.  IBS is based on the position of the day’s close in relation to the day’s range: it takes a value of 0 if the closing price is the lowest price of the day, and 1 if the closing price is the highest price of the day.

More formally:

IBS  =  (Close – Low) / (High – Low)

The IBS effect may be related to intraday over-reaction to news or market movements, which are then ”corrected” the next day.  It serves as a measure of the tendency of a price series to mean-revert over daily horizons.  I use the term “daily” advisedly: so far as I am aware, there has been no research (including my own) demonstrating the existence of an IBS effect at time horizons shorter, or longer, than one day.  Indeed, there has been very little in the way of academic research into the concept of any kind, which is strange considering how compelling are the results it is capable of producing.  Practitioners have been happy enough with that state of affairs, content to deploy this neglected indicator in their trading strategies, where it has often proved to be extremely useful (we use IBS in one of our volatility strategies). Since 2013, however, the cat has been let out of the bag, thanks to an excellent research paper by Alexander Pagonidis, who writes an interesting quantitative finance blog.

The essence of the idea is that stocks that close in the lowest part of the daily range, with an IBS of below, say, 0.2, will tend to rally the next day, while stocks that close in the highest quintile will often decline in value in the following session.  In his paper “The IBS Effect: Mean Reversion in Equity ETFs” (2013), Pagonidis researches the IBS effect in equity index ETFs in the US and several international markets.  He confirms that low IBS values in these assets are associated with high returns in the following day session, while high IBS values are associated with low returns. Average returns when IBS is below 0.20 are .35% ,while average returns when IBS is above 0.80 are -0.13%. According to his research, this effect has been present in equity ETFs since the early 90s and has been highly consistent through time.

SSALGOTRADING AD

IBS Strategy Performance

To give the reader some idea of the potential of the IBS effect, I have reproduced below equity curves for the IBS strategy for the SPDR S&P 500 ETF Trust (SPY) and iShares MSCI Singapore ETF (EWS) index ETFs over the period from 1999 to 2016.  The strategy buys at the close when IBS is below 0.2, and sells at the close when IBS exceeds 0.8, liquidating the position at the following market close. Strategy CAGR over the period has been of the order of 13% for SPY and as high as 40% for EWS, ignoring transaction costs.

IBS Strategy Chart SPY EWS

 

Note that in both cases strategy returns for SPY and EWS have diminished in recent years, turning negative in 2015 and 2016 YTD and this is true for ETFs in general.  It remains to be seen whether this deterioration in strategy performance is temporary or permanent.  There are some indications that the latter holds true, but the evidence is not quite definitive.  For example, the chart below shows daily equity curve for the SPY IBS strategy, with 95% confidence intervals for the latest 100 trades (up to the end of May 2016), constructed using Monte-Carlo bootstrap.  The equity curve appears to have penetrated the lower bound, indicating a statistically significant deterioration in the performance of the IBS strategy for SPY over the last year or so (EWS is similar).  That said, the equity curve does fall inside the boundaries of the 99% confidence interval, so those looking for greater certainty about the possible breakdown of the effect will need to wait a little longer for confirmation.

 

SPY IBS MSA

 

Whatever the outcome may be for SPY and other ETFs going forward, it is certainly true that IBS effects persist strongly for some individual equities, Exxon-Mobil Corp. (XOM) being a case in point (see below).  It’s worth taking note of the exceptional performance of the XOM IBS strategy during the latter quarter of 2008.  I will have much more to say on the application of the IBS indicator for individual equities in a future blog post.

 

XOM IBS Strategy

 

The Role of Range, Volume, Bull/Bear Markets, Volatility and Seasonality

Pagonidis goes on to detail several further important findings in relation to IBS.  It is clear from his research that high volatility is related to increased predictability of returns and a more powerful IBS effect, in particular the high IBS-negative return aspect.  As might be expected, the effect is also larger after days with high range, both for high and low IBS extremes.

Volume turns out to be especially important for  U.S. index ETFs:  in fact, the IBS effect only appears to work on high-volume days.

Pagonidis also separates the data into bull and bear market environments, based on whether 200-day returns are positive or not.  The size of the effect is roughly similar in each environment (slightly larger in bear markets), but it is greater in the direction of the overall trend: high IBS readings are followed by larger negative returns during bear markets, and vice versa.

Day of Week Effect

The IBS effect is also strongly seasonal, having the greatest impact on returns from Monday’s close to Tuesday’s close, as illustrated for the SPY ETF in the chart below.  This accounts for the phenomenon known popularly as “Turnaround Tuesday”, i.e. the tendency for the market to recover strongly from losses on a Monday.  The day-of-week effect is weakest for Fridays.

 

SPY DOW

 

The mean of the returns distribution is not the only aspect that IBS can predict. Skewness also varies significantly between IBS buckets, with low IBS readings being followed by highly skewed returns, and vice versa. Close-to-close returns after a bottom-bucket IBS day have average skewness of 0.65 across Equity Index ETF products, while top-bucket IBS days are followed by returns with skewness of 0.03. This finding has very useful risk management applications for investors concerned with tail risk.

IBS as a Filter for a Swing Trading Strategy in QQQ

The returns to an IBS-only strategy are both statistically and economically significant. However, commissions will greatly decrease the returns and increase the maximum drawdowns, however, making such an approach challenging in the real world. One alternative is to combine the IBS effect with mean reversion on longer timescales and only take trades when they align.

Pagonidis offers a simple demonstration using the Cutler’s RSI indicator that shows how the IBS effect can be used to boost returns of a swing trading strategy while significantly decreasing the number of trades needed.

Cutler’s RSI at time t is calculated as follows:

 

RSI

 

Pagonidis tests a simple, long-only strategy that trades the PowerShares QQQ Trust, Series 1 (QQQ) ETF using the Cutler’s RSI(3) indicator:

• Go long at the close if RSI(3) < 10

• Maintain the position while RSI(3) ≤ 40

 filter these returns by adding an additional rule based on the value of IBS:

• Enter or maintain long position only if IBS ≤ 0.5

Pangonis claims that the strategy produces rather promising results that “easily beats commissions”;  however, my own rendition of the strategy, assuming commissions of $0.005 per share and slippage of a further $0.02 per share produces results that are distinctly less encouraging:

EC0

 

Pef0

Strategy Code

For those interested, the code is as follows:

Inputs:
RSILen(3),
RSI_Entry(10),
RSI_Exit(40),
IBS_Threshold(0.5),
Initial_Capital(100000);
Vars:
nShares(100),
RSIval(0),
IBS(0);
RSIval=RSI(C,RSILen);
IBS = (C-L)/(H-L);

nShares = Round(Initial_Capital / Close,0);

If Marketposition = 0 and RSIval > RSI_Entry and IBS < IBS_Threshold then begin
Buy nShares contracts next bar at market;
end;
If Marketposition > 0 and ((RSIval > RSI_Exit) or (IBS_Threshold > IBS_Threshold)) then begin
Sell next bar at market;
end;

Strategy Optimization and Robustness Testing

One can further improve performance by optimizing the trading system parameters, using Tradestation’s excellent Walk Forward Optimization (WFO) module.  This allows us to examine the effect of re-calibrating the strategy parameters are regular intervals, testing the optimized model on out-of-sample data sets of various sizes.  WFO can be used, not only optimize a strategy, but also to examine the sensitivity of its performance to changes in the levels of key parameters.  For example, in the case of the QQQ swing trading strategy, we find that profitability increases monotonically with the length of the RSI indicator, and this effect is especially marked when an IBS threshold level of 0.2 is used:

Sensitivity

 

Likewise we can test the consistency of the day-of-the-week effect over several OS data sets of  varying size and these tests are consistent with the pattern seen earlier for the IBS indicator, confirming its role as a filter rule in enhancing system profitability:

Distribution Analysis

 

A model that is regularly re-calibrated using WFO is subjected to a series of tests designed to ensure its robustness and consistency in live trading.   The tests include the following:

 

WFO

 

In order to achieve an overall pass rating, the system is required to pass all five tests of its out-of-sample performance, from which Tradestation deems it likely that the system will continue to perform well in live trading.  The results from this procedure appear much more promising than the strategy in its original form, as can be seen from the performance table and equity curve chart shown below.

EC1

Perf1

 

However, these results include both in-sample and out-of-sample periods.  An examination of the results from the WFO indicate that the overall efficiency of the strategy is around 55%, meaning that the P&L produced by the system in out-of-sample periods amounts to a little over one half of the rate of profit produced during in-sample periods.  Going forward, therefore, we might expect the performance of the system in live trading to be only around half as good as shown here.  While this is still superior to the original system, it may not be considered good enough.  Nonetheless, for the purpose of illustrating the benefits of the IBS indicator as a trade filter, it makes the point.

Another interesting example of an IBS-based trading strategy in the QQQ and SPY ETFs can be found in the following blog post.

Conclusion

Internal Bar Strength is a powerful mean-reversion indicator for equity products traded at daily frequencies, with a consistent effect that has continued from the 1990s through to the current decade. IBS can be used on its own in mean-reversion strategies that have worked well for both US equities and US and International equity index ETFs, or used as a trade filter when combined with other alpha signals.

While there is evidence of a weakening of the IBS effect since around 2013 this is not yet confirmed statistically (at the 99% confidence level) and may simply be the result of normal statistical variation in its efficacy.

 

 

Combining Momentum and Mean Reversion Strategies

The Fama-French World

For many years now the “gold standard” in factor models has been the 1996 Fama-French 3-factor model: Fig 1
Here r is the portfolio’s expected rate of return, Rf is the risk-free return rate, and Km is the return of the market portfolio. The “three factor” β is analogous to the classical β but not equal to it, since there are now two additional factors to do some of the work. SMB stands for “Small [market capitalization] Minus Big” and HML for “High [book-to-market ratio] Minus Low”; they measure the historic excess returns of small caps over big caps and of value stocks over growth stocks. These factors are calculated with combinations of portfolios composed by ranked stocks (BtM ranking, Cap ranking) and available historical market data. The Fama–French three-factor model explains over 90% of the diversified portfolios in-sample returns, compared with the average 70% given by the standard CAPM model.

The 3-factor model can also capture the reversal of long-term returns documented by DeBondt and Thaler (1985), who noted that extreme price movements over long formation periods were followed by movements in the opposite direction. (Alpha Architect has several interesting posts on the subject, including this one).

Fama and French say the 3-factor model can account for this. Long-term losers tend to have positive HML slopes and higher future average returns. Conversely, long-term winners tend to be strong stocks that have negative slopes on HML and low future returns. Fama and French argue that DeBondt and Thaler are just loading on the HML factor.

SSALGOTRADING AD

Enter Momentum

While many anomalies disappear under  tests, shorter term momentum effects (formation periods ~1 year) appear robust. Carhart (1997) constructs his 4-factor model by using FF 3-factor model plus an additional momentum factor. He shows that his 4-factor model with MOM substantially improves the average pricing errors of the CAPM and the 3-factor model. After his work, the standard factors of asset pricing model are now commonly recognized as Value, Size and Momentum.

 Combining Momentum and Mean Reversion

In a recent post, Alpha Architect looks as some possibilities for combining momentum and mean reversion strategies.  They examine all firms above the NYSE 40th percentile for market-cap (currently around $1.8 billion) to avoid weird empirical effects associated with micro/small cap stocks. The portfolios are formed at a monthly frequency with the following 2 variables:

  1. Momentum = Total return over the past twelve months (ignoring the last month)
  2. Value = EBIT/(Total Enterprise Value)

They form the simple Value and Momentum portfolios as follows:

  1. EBIT VW = Highest decile of firms ranked on Value (EBIT/TEV). Portfolio is value-weighted.
  2. MOM VW = Highest decile of firms ranked on Momentum. Portfolio is value-weighted.
  3. Universe VW = Value-weight returns to the universe of firms.
  4. SP500 = S&P 500 Total return

The results show that the top decile of Value and Momentum outperformed the index over the past 50 years.  The Momentum strategy has stronger returns than value, on average, but much higher volatility and drawdowns. On a risk-adjusted basis they perform similarly. Fig 2   The researchers then form the following four portfolios:

  1. EBIT VW = Highest decile of firms ranked on Value (EBIT/TEV). Portfolio is value-weighted.
  2. MOM VW = Highest decile of firms ranked on Momentum. Portfolio is value-weighted.
  3. COMBO VW = Rank firms independently on both Value and Momentum.  Add the two rankings together. Select the highest decile of firms ranked on the combined rankings. Portfolio is value-weighted.
  4. 50% EBIT/ 50% MOM VW = Each month, invest 50% in the EBIT VW portfolio, and 50% in the MOM VW portfolio. Portfolio is value-weighted.

With the following results:

Fig 3 The main takeaways are:

  • The combined ranked portfolio outperforms the index over the same time period.
  • However, the combination portfolio performs worse than a 50% allocation to Value and a 50% allocation to Momentum.

A More Sophisticated Model

Yangru Wu of Rutgers has been doing interesting work in this area over the last 15 years, or more. His 2005 paper (with Ronald Balvers), Momentum and mean reversion across national equity markets, considers joint momentum and mean-reversion effects and allows for complex interactions between them. Their model is of the form Fig 4 where the excess return for country i (relative to the global equity portfolio) is represented by a combination of mean-reversion and autoregressive (momentum) terms. Balvers and Wu  find that combination momentum-contrarian strategies, used to select from among 18 developed equity markets at a monthly frequency, outperform both pure momentum and pure mean-reversion strategies. The results continue to hold after corrections for factor sensitivities and transaction costs. The researchers confirm that momentum and mean reversion occur in the same assets. So in establishing the strength and duration of the momentum and mean reversion effects it becomes important to control for each factor’s effect on the other. The momentum and mean reversion effects exhibit a strong negative correlation of 35%. Accordingly, controlling for momentum accelerates the mean reversion process, and controlling for mean reversion may extend the momentum effect.

 Momentum, Mean Reversion and Volatility

The presence of  strong momentum and mean reversion in volatility processes provides a rationale for the kind of volatility strategy that we trade at Systematic Strategies.  One  sophisticated model is the Range Based EGARCH model of  Alizadeh, Brandt, and Diebold (2002) .  The model posits a two-factor volatility process in which a short term, transient volatility process mean-reverts to a stochastic long term mean process, which may exhibit momentum, or long memory effects  (details here).

In our volatility strategy we model mean reversion and momentum effects derived from the level of short and long term volatility-of-volatility, as well as the forward volatility curve. These are applied to volatility ETFs, including levered ETF products, where convexity effects are also important.  Mean reversion is a well understood phenomenon in volatility, as, too, is the yield roll in volatility futures (which also impacts ETF products like VXX and XIV).

Momentum effects are perhaps less well researched in this context, but our research shows them to be extremely important.  By way of illustration, in the chart below I have isolated the (gross) returns generated by one of the momentum factors in our model.

Fig 6