Author: byran

SA equities: Momentum continues to be crushed

SA equities: Momentum continues to be crushed

The markets have definitely changed. The low volatility, easy monetary policy and search for yield proved to be a boon for momentum stocks. Going long high momentum and short low momentum stocks had almost become a staple trade for many. Momentum, however, has been crushed in 2016 and in turn that trade has now reversed.

high-low-momentum

In South Africa the resurgence in value stocks started with the strong run up in resource stocks, however, it would be wrong to assume that the rise in value (and fall in momentum) is only due to resource stocks. In our value portfolios the bulk of our allocation was actually within the industrial sector and we, surprisingly, held few resource counters. This view is backed up by the recent article in Moneyweb; Value vs growth – are we at an inflection point? . Cannon Asset Managers’ Super Dogs portfolio has no resources and yet managed to be up 8% YTD against a FINDI Index which is down c.7%.

Where I disagree with the article is whether we actually are at an inflection point. No. We’re way past that point. The value strategy has started to work again (at least in South Africa) and has been outperforming the market for the past 12 months at least. The stocks that banked you outperformance in the past with their low volatility, high momentum characteristics are just not cutting it anymore.

momentum-value

So where to from here? My view is that value continues to outperform, probably at a lower rate as more and more investors get hurt by a tilt towards a high momentum strategy and start switching to value.

Momentum strategies do not like volatility. Value strategies on the other hand love volatility. And volatility is on the rise.

momentum-vol

So to are short-term rates (although who knows for certain these days) and historically value strategies have outperformed the market when short-term rates rise.

Our third reason for favouring a continued run in value stocks is a model we have put together which attempts to estimate the return of a high momentum strategy vs. a high value strategy over the next 3 months. It’s strongly negative, indicating outperformance by a high value strategy over a high momentum strategy.

Momentum relative return

Although every effort is made to ensure accuracy in the data and analysis presented on this site, this cannot be guaranteed. Nothing on this site constitutes investment advice, nor should it be taken as such. And obviously, past results do not guarantee future performance.

SA equities: Basic fundamental factor portfolios

SA equities: Basic fundamental factor portfolios

I wanted to switch gears a bit from momentum and look at some fundamental factor screening portfolios. Here I am only interested in using historical fundamental data. Also note that I have lagged all fundamental data by 3 months to reduce the impact of backfilling.

I have tried to use fairly basic factors. A few “value” based factors:

  • Dividend Yield
  • EV/EBITDA
  • Price-to-cashflow
  • Price-to-book
  • Price-to-earnings

and what I guess you would call “quality” based factors:

  • Return on assets
  • Return on equity
  • Return on invested capital

There are a lot more factors but I just wanted to get started on the simple ones.

(As an aside here is a nice post on the impact of using forward or trailing PE ratios: https://blogs.cfainstitute.org/investor/2016/07/12/dumb-alpha-trailing-or-forward-earnings/)

In general I am keeping the portfolio construction process very similar to the momentum portfolio construction process from previous posts.

  • At the end of each month I rank the top 80 stocks (as at each month) according to a specific fundamental factor.
  • Place each stock in one of five portfolios, from most favourable to least. For example from highest DY to lowest or lowest PE to highest.
  • Equal weight each stock using closing prices of the following trading day (i.e. I am assuming that the portfolio is readjusted at the end of the next day).
  • Allow the weights to float until the next formation/month (this replicates an actual portfolio – each stock is equal weighted at the beginning of the month but price changes will change the weights during the month).
  • Returns are net of dividends and capital adjustments and exclude transaction costs.

For every factor I just going to put a chart of the cumulative performance and then a statistics table. I provide a brief explanation of the table below, otherwise I also provide some of my thoughts on each factor. In the next post I will look at combining the value factors with the quality factors and see if we can improve the portfolios.

Just in general though, a few things I am looking for:

  • Good separation of portfolios in a sensible order should illustrate selection power of the factor (i.e. lowest PE should have the best return and declining from there as we move down the quintiles)
  • A Sharpe of over 1 and a Sortino of close to 1.5
  • A good size alpha per unit maximum drawdown. For me this is a measure of “easy” it is to stick with the strategy in a down turn
  • Obviously a large probability of a stock leaving the portfolio with a positive return
  • A fair time in the portfolio – we want to limit churn as much as possible.

The chart gives you the cumulative performance (log-scale) for each quintile portfolio. I have also provided the performance of the market. However, note this isn’t a JSE index. It’s a market cap portfolio which is rebalanced every month and will differ slightly from official indices due to, amongst other reasons, free float weights and the quarterly rebalance (as opposed to monthly) of JSE indices.

In the table I have shown the annualised return, volatility and downside deviation. Also the corresponding Sortino and Sharpe ratios. I have also looked at maximum drawdowns and a measure I call alpha per unit maximum drawdown (Alpha_maxDD). This measures the alpha (over the market cap index) relative to the strategy’s maximum drawdown. The higher the alpha per until maximum drawdown the better. Prob_Up is the probability that a stock leaving the portfolio does so with a positive return since entering and Avg_Time is the average time a stock will stay in a portfolio (measured in months).

 


Dividend Yield (DY)

DY - Chart

 

DY doesn’t fair too badly although the spread in returns is not that great and quintile 2 has just recently surpassed quintile 1. Quintile 4 also sits in a weird spot between quintile 2 and the market.

DY - Table


EV/EBITDA

EVEBITDA - Chart

EV/EBITDA looks promising. A nice sensible spread in returns: quintile 1 and 2 with the highest returns, quintile 4 and 5 with the lowest. Although. Quintile 5 has caught up a lot of ground recently while quintiles 1 and 2 have pulled back significantly. Probably a sign of the period we have just emerged from. Important to note financial stocks will not have an EV/EBITDA and so will naturally be excluded. (Perhaps we can use PCF in the case of financials and EV/EBITDA for everything else)

EVEBITDA - Table


Price-to-book (PB)

 PB - Chart

A favourite in the market, PB, has probably the worst experience. Both quintile 4 and 5 outperform the other portfolios* (i.e. the highest and next highest PB stocks). Every other portfolio has returns similar to the market. Not a great separator then.

*Perhaps there is something wrong in my back test but I have triple checked this!

PB - Table


Price-to-cashflow (PCF)

 PCF - Chart

PCF is a bit of a weird one. Quintile 2 and 3 are clear winners while quintile 1 is sort of middle of the pack. Just as in the PE case (see below), avoiding the highest PCF stocks seems to be a good way to go.

PCF - Table


Price-to-Earnings (PE)

 PE - Chart

The fan favourite, the PE ratio. It had a great run until about 2013 where things went horribly wrong (probably due to resource heavy portfolios). Quintile 4 actually does really well – almost as well as quintile 1 – which makes me question its ability to separate stocks accurately but this is only recently. Pre-2009 wasn’t much difference between the portfolios. It does seem to indicate that you should avoid the highest PE stocks at all costs.

PE - Table


Return on Assets (ROA)

 ROA - Chart

ROA actually comes across as the best factor. A decent and stable spread in the portfolio returns. Quintile 1 returns of close to 20% per annum, alpha per unit max drawdown of over 12% and an average time spent in the portfolio of 11 months. My pick of this bunch really.

ROA - Table


Return on Equity (ROE)

 ROE - Chart

ROE is another favourite which actually doesn’t look too great. Quintile 2 is the second worst performing portfolio while the second best portfolio happens to be quintile 4. Not really the order we are looking for.

ROE - Table


Return on Invested Capital (ROIC)

ROIC - Chart

ROIC is actually fairly decent. Not as good as ROA but better than ROE. There is some order in the returns but quintile 5 just does too well for me to consider it a good stock separator.

ROIC - Table

 

Although every effort is made to ensure accuracy in the data and analysis presented on this site, this cannot be guaranteed. Nothing on this site constitutes investment advice, nor should it be taken as such. And obviously, past results do not guarantee future performance.

Momentum in SA equities over different lookback periods

Momentum in SA equities over different lookback periods

It’s been a while since a posted the previous post in which we looked at momentum in SA equities over a 6 month look back period (skipping the latest month). We also showed how the volatility of the highest stock momentum portfolio looked reasonably predictable (or at least wasn’t random noise). Is 6 month’s optimal though for a lookback period? That’s something we want to address in this post.

In general we keep the portfolio construction process the same as in the previous post:

  • At the end of each month calculate the momentum (over a specific lookback period) of the top 80 stocks (skipping the most recent month).
  • Rank the stocks from highest to lowest momentum placing each stock in one of five portfolios based on this ranking (highest to lowest).
  • Equal weight each stock using closing prices of the following trading day (i.e. we are assuming the we readjust the portfolio at the end of the next day).
  • Allow the weights to float until the next formation/month (this replicates an actual portfolio – each stock is equal weighted at the beginning of the month but price changes will change the weights during the month).
  • Returns are net of dividends and capital adjustments and exclude transaction costs.

 

Going long the top momentum stocks

First we look at portfolios that go long the top quintile of momentum stocks (equal weighted and re-balanced at the end of each month). Below are the cumulative returns (on a log-scale):

Momentum over different lookbacks - long only

It’s a bit messy but we have highlighted the top and bottom two.  Interestingly 6 months’ is a rather “middle-of-the-pack” performer. The more longer-term, 12 and 24 month portfolios seem to do better (and consistently so). As expected the longer the lookback period the lower the returns, although too short and the returns are even worse as highlighted by the 2 month portfolio.

We have shown the total return (annualised) and Sharpe ratios below. This further illustrates this “polynomial” relationship – if the lookback period is either too short or too long a (long-only) momentum portfolio’s return degrades. The 12 to 24 month period again seems to be fairly optimal.

Long only - share and cagr

 

A long-short portfolio

The above analysis was only for a long-only portfolio (long the top momentum stocks). Here we look at a long-short portfolio: long the top momentum stocks and short the lowest momentum stocks. This will give us a good idea of how sustainable momentum is at different lookback periods and whether mean reversion is more prevalent at specific periods.

 

Momentum over different lookbacks - long-short

We have to be careful here, since the period above only stretches from 2004. Momentum worked really well since 2011 – so no matter which lookback period we used we would have done well. Clearly this isn’t going to continue forever and so we need to contrast post-2011 performance with pre-2011 performance.

The 36 and 48 month portfolios seem to be the worst performers and returns are actually negative until 2011. The best performer is once again the 12 month portfolio (even pre-2011).

The annualised returns and Sharpe ratios highlight a similar relationship to the lookback period, although the 6 – 12 month portfolios seem to be fairly optimal. Returns decline from the 12 month lookback onwards.

Long short - share and cagr

Conclusion

You can never be sure what the future will hold in terms of optimisations and how these parameters will evolve going forward. However, from both the long and long/short portfolios it seems like 12 months has done pretty well both pre- and post-2011. However, from the charts of the annualised returns and Sharpe ratios it seems like anything between 6 and 12 months is a decent pick as a lookback period.

 

Although every effort is made to ensure accuracy in the data and analysis presented on this site, this cannot be guaranteed. Nothing on this site constitutes investment advice, nor should it be taken as such. And obviously, past results do not guarantee future performance.

Momentum in South African equities

Momentum in South African equities

Momentum has proven to be one of the most pervasive factors out there in the investing/trading world. A number of academic papers (beginning with the seminal Jegadeesh and Titman paper) have shown momentum to be a persistent and somewhat unexplained (at least by traditional market theory) factor. In South Africa there have been a few academic papers on the subject. A comprehensive paper by Muller and Ward, for example, looked at a variety of factors and found 12 month momentum to be one of the strongest (amongst a few others). They found a momentum portfolio formed on the prior 12 month returns and held for 3 months outperformed the ALSI by roughly 9% per annum over the period (1984-2012).

Lets take a closer look at momentum portfolios, specifically in the South African equity space. We start off by narrowing down the universe to the top 80 stocks on the JSE (data from Bloomberg) from 1 Jan 2000 to 22 March 2016. That is, in our analysis, the top 80 stocks as at the end of each month are taken as our universe with everything else excluded (for that period). We then form 5 portfolios based on 6 month momentum (we will play around with the formation period in a later post). We also exclude the most recent month when calculating momentum due to some evidence that strong momentum stocks exhibit a reversal effect in following next month (this was found in the original Jegadeesh and Titman paper and further elaborated on in a follow up paper – the impact has sometimes been found to be rather marginal).

To summarise:

  • At the end of each month calculate the (6 month) momentum of the top 80 stocks (skipping the most recent month).
  • Rank the stocks from highest to lowest momentum placing each stock in one of five portfolios based on this ranking (highest to lowest).
  • Equal weight each stock using closing prices of the following trading day (i.e. we are assuming the we readjust the portfolio at the end of the next day).
  • Allow the weights to float until the next formation/month (this replicates an actual portfolio – each stock is equal weighted at the beginning of the month but price changes will change the weights during the month).
  • Returns are net of dividends and capital adjustments and exclude transaction costs.

 

Momentum portfolio returns using 6 month momentum

We show the returns for each quintile (i.e top 8 stocks by momentum in quintile 1, and so on). There appears to be quite a good separation – especially between quintile 1 (high momentum) and quintile 5 (low momentum).

Momentum quintiles (6m)

 

Below we show the relative returns of the highest momentum and lowest momentum stocks which highlights the long-term success of momentum. This equates to roughly a 22% outperformance of the low momentum portfolio per annum on average (by the high momentum portfolio).

Q1-Q5 (6m)

When does high momentum underperform?

Most of this outperformance is actually only due to two periods: 2001 – 2007 and 2012 – 2015. High momentum actually underperformed low momentum by roughly 30% during the initial stages of the recovery (2009/2010). The underperformance during the initial stages of the recovery in 2009 is an effect described in the paper Momentum Crashes (Daniel & Moskowitz).

They find that the loser (low momentum) portfolio has much higher beta following a large market decline. Therefore, if you go long high momentum and short low momentum stocks you end up with a large negative beta position (high momentum portfolio’s beta minus a much larger low momentum portfolio’s beta). This means as the market recovers as a whole, higher beta stocks (which are more likely to be in the low momentum portfolio) will do better (thereby outperforming the high momentum portfolio). Daniel & Moskowitz also explain the long-short momentum portfolio as a written call option on the market just after a large decline (due to the large negative beta of the long-short portfolio). Therefore, as the market recovers the call option goes into the money and you (as the writer of the call option) start to accrue losses.

The analogy to written call options is useful, because with a written call option you become short volatility. That is, as volatility increases (all else being equal) the value of the call option rises (and consequently your losses increase). This makes sense if you think about the market conditions which a momentum strategy would thrive in or conversely in which it woudl struggle. A momentum strategy would do struggle with whipsaws, which are present during heightened volatility (hence the whipsaw). A momentum strategy would do well when a stock grinds higher (or lower) at a consistent pace (i.e. when experiencing low volatility).

Daniel & Moskowitz find there is an inverse relationship between volatility and the long-short portfolio’s returns, but only during bear markets. More importantly though, they find that the volatility of the winners portfolio is relatively predictable and using this they are able to scale in and out of the momentum strategy and improve returns and Sharpe ratios.

 

Rolling 6M volatility

In the chart above we plot the rolling 6 month volatility of the high momentum (quintile 1) portfolio for SA equities. The volatility does seem to follow a pattern that reverts to some short-run mean over the short-term and some long-run mean in the long-term – something we may be able to model.

The role of volatility in relative performance

Below we show the short-run mean against the relative return of high momentum against low momentum (long high, short low momentum). Keeping in mind that there is some lag in the smoothed rolling 6M volatility, there does appear to be some relationship. Over the long-term, as volatility increased from 2003 to 2008, the relative return declines until eventually the high momentum portfolio underperforms the low momentum portfolio over 2009-2010.

Volatility vs 12M Rel Returns

Over the period 2011-2014 volatility has declined significantly to near all-time lows, benefiting high momentum. However, of late, markets have experienced heightened volatility, something momentum strategies do not enjoy. The key question is: will these conditions endure and see momentum strategies reverse gains? Momentum has not done well YTD (at least in South Africa) as evident from the sharp decline in relative returns in the chart above. Perhaps volatility declines in the short-term, but it seems as though in the long-term volatility may increase – something to keep in mind when dealing with momentum strategies.

Conclusion

We are, arguably, in the midst of a regime change. We have experienced low interest rates and “easy money” but this era is likely at an end. We should therefore be wary of strategies that have worked over the past five years and whether these will continue in the future.

Momentum as a strategy seems to have done well in South African equities over the past 15 years and more recently from 2012 – 2015. Should we expect this to continue? To address this we have looked at the role volatility has played in momentum strategy returns. Both in academic research and in our own analysis we have found volatility often has an inverse relationship with momentum strategy returns.

The volatility of the highest momentum portfolio does appear to follow some structure – reverting to short-run mean in the short-term and a long-run mean in the long-term. Volatility may decline in the short-term, but it does appear that long-term volatility is on the rise and this is likely to prove a headwind to any momentum strategies.

Some additional questions and points we may want to address in the future:

  • The performance of momentum over lags other than six months and over different holding periods
  • The impact of skipping the most recent month
  • Can we model the volatility of the past winners portfolio?
  • What are the implications for a market cap index (seeing as this is merely a long-only momentum strategy)?

 

Although every effort is made to ensure accuracy in the data and analysis presented on this site, this cannot be guaranteed. Nothing on this site constitutes investment advice, nor should it be taken as such. And obviously, past results do not guarantee future performance.