Labels. On clothes, they are a shortcut to understanding everything of significance. Where and from what it was made, its dimensions, and most importantly, what it should and should not be exposed to in its ongoing wear and maintenance. For instance, did you know the five icons on the above label mean: Wash at below 40 C, Do not bleach, Do not tumble dry, Iron at medium temperature (two dots for medium) and Do not dry clean.
On investors too, labels are a short-cut to understanding everything of significance. Where they come from, time they have spent being an investor, and most importantly, where they will and will not invest. What would the icons on an imagined “Value Investor” label look like? Maybe: Invest in “quality” only, Do not invest at historic highs, Do not expose to financial media, Quote Buffett liberally and Moats!
Those with the instincts of editors might point out the inconsistency. That the clothes’ labels are meant for the people using them while the investor labels seem to be for themselves.
Let’s just call it the investment-blogger license and leave it at that.
The labels above might have given you the impression that I’m trolling value investors and that I’m a chart-laden technical trader, algorithmically switching in and out of positions faster than you can say “MACD Bullish Crossover”. Far from it. The first investment book not prescribed as part of any Finance course I read was ‘The Intelligent Investor’, and the idea of the stock price as something tethered to the value of the underlying business, sometimes tightly, other times loosely just made sense. Since then I have been a staunch believer in the idea of reversion to the mean, gravitating towards out-of-favour sectors and companies, happy to accumulate positions in companies I felt had sound balance sheets but short-term challenges which are being viewed pessimistically by the market.
This investment philosophy has worked reasonably well for me over the twelve or so years I’ve been investing with intent, an ongoing process of adding/improving frameworks and tools to better define what is quality and value.
Challenging what you know
“Your assumptions are your windows on the world. Scrub them off once in a while, or the light won’t come in” – Isaac Asimov
Since what we don’t know about markets is so many orders of magnitude larger than what we do know, it’s self-sabotage to refuse to even consider other schools of thought about what works in the markets. While it’s unlikely that I’ll be embarking on astrological stock-picking any time soon, I have been intrigued enough by momentum investing for a while, that I have been conducting an experiment with it, with actual money.
The rest of this post is about what I’ve observed from the experiment. It should be evident from the title of the post, that this is in no way an expert primer on momentum investing but a learn from first-principles approach that is still evolving.
The (bare) minimum. Momentum Investing is a system of buying and selling stocks based on recent returns. Momentum investors buy outperforming securities, and avoid, or short underperforming securities. No balance sheets, profit & loss or cash flow statements. This school of thought assumes all relevant information is already embedded in the stock price. And outperforming securities will continue to outperform in absence of significant headwinds.
The base case for momentum
A recent Morgan Stanley report makes a sensational claim about the superiority of momentum investing over every other strategy.
Since they did not explain the chart or the methodology used to classify stocks into individual style buckets, I’d take this chart with a large pinch of salt (“junk” outperforming growth, value and quality?)
I won’t get into them here but there is other more credible research that empirically suggests the persistence of momentum as a driver of alpha i.e. returns above the broader market.
A simple, quick and dirty backtest
Let’s backtest the most simplistic momentum strategy possible. A portfolio of NSE 500 stocks with the 10 highest returns over the last 12 months and rebalance monthly.
The test runs from March 2014 to April 2018, 49 months. Why not longer or shorter? It’s random. As a value investor, I have no reason to cherry pick the timeframe to make it look good. Heck, a small part of me wanted to be able to say that momentum is hogwash.
The result. Momentum vs NSE 500.
The Naive Momentum 10 (NM10) portfolio simply buys 10 stocks with highest price appreciation over the last 12 months, holds until the last trading day of the month, and repeats the process each month.Assuming you start March 2014 with ₹100 in the NM10 portfolio, and reinvest gains into the portfolio, you end on 18th April 2018 with ₹265, XIRR of 27%. In comparison, the NSE 500 would be worth ₹157 at a relatively modest XIRR of 12%.
A simple visual inspection tells you the NM10 would’ve given you a few sleepless nights over that time period.
This is confirmed by comparing drawdowns from peak in both cases that shows much bigger swings in value for NM10, as much as 35% down from peak values.
Another way to look at the data is to compare monthly performance over the four years instead of cumulatively.
Rubber, meet Road
At this point my options were:
- Dismiss the results as pure randomness and move on
- Apply more rigorous conditions (this backtest suffers from survivorship bias because it only considered stocks currently in the NSE 500)
- Try it for real
Unless you’re getting paid as a tenured academic for writing research papers, the objective of any backtest should be to decide whether it gives you enough confidence to try it in the real world.
You guessed it. I picked option 3 and selected my first ever set of momentum stocks in late Sep 2017. I allocated ~5% of my portfolio to this strategy, enough to keep me interested, not enough for me to lose sleep.
Since then it has performed reasonably in what has been a volatile market.
The April 2018 Portfolio for this strategy looked like this:
An equal-weighted portfolio of these 10 stocks returned 15% for the month. You could cry foul that the bulk of those came from one stock, Indiabulls Ventures but that’s a consistent pattern over the period that I have implemented this strategy where one of two stocks deliver the bulk of the returns, a couple correct and the rest move marginally.
Over a couple of iterations I have applied two additional criteria:
- On ‘Quality’ of momentum derived from the consistency of upward movement which is just a fancier way of saying “number of up days as % of total days”
- On recency of price momentum i.e. to also take into account price movement in the last three and six months
My work-in-progress takeaway from my experiment is that there is some information in price momentum. Whether it’s because of insiders who know more than us, or just a self-fulfilling prophecy set in motion by momentum traders, I do not know. But the results so far justify a small component of my portfolio to be deployed in it. And because momentum seems the antithesis of contrarian value buying, it offers some much-needed strategy diversification to my portfolio. Whether it works or bombs, I will do a follow up post on how NM10 2.0 does over the next few months.
Note: Capitalmind Premium has a momentum portfolio too – running since 2013 in different avatars. In 2017, it did a relatively smart 65%+ return, and has had changes only once a month. Click here to subscribe!