As of 9th May 2020: Gold (+48%) has outperformed Nifty (-19%) by 67%. In one year.
You can not time the market. But wait…
Imagine if you had known this chart back in May 2019. Heck, back in Feb 2020 would’ve worked too.
But you can not. Because timing the market is impossible. Which is why you should buy (or SIP) and hold. End of story?
To start with, let’s keep it simple and not pretend that we could have predicted Gold’s performance. But could we have gotten out of the NIFTY before it endured all that damage?
Over the last one year, the timed-NIFTY strategy (suspend your rightful disbelief for a moment) apparently manages to retain almost all its value. Compared to that, someone who had bought the NIFTY a year ago would be down 19%.
The timing strategy would have moved to cash in the first week of March 2020 and would not have re-entered as of May 2020.
Your inner skeptic is smirking, like when you receive those spam “can’t lose trading idea” SMS’s with company names like MicraSoft Technologies.
Except, maybe consider that the very same folks who say “of course you can’t time the market”, often in the same breath say, “but here’s a set of stocks / sectors / asset classes that will beat the market”.
Different phrases that mean the same thing i.e. do better than the market. That should at least have you wondering. Is it really that ridiculous?
The Simple Moving Average Exit-Entry
This is the simple strategy we applied that resulted in the one-year chart above. To exit the NIFTY every time it crosses below a pre-determined moving average threshold, and to re-enter only when the NIFTY crosses above that threshold. No macro analysis of what’s happening with global liquidity, or what central banks might or might not do. Just the price.
Here’s the outcome of such a strategy going back 16 years, to 2005.
Cumulative line charts on arithmetic scales don’t offer a clear idea of whether the outperformance is purely a result of starting point bias and one or two great years.
To get a better sense of how two strategies compare, we need to look at standalone annual returns and compare drawdowns from peak.
Buy & Hold outperforms Rule-based timing in 12 out of 16 years. What’s all the fuss about then?
The magic happens in the years that the rule-based timing strategy outperforms Buy & Hold. 2008. 2011. And now. Timing rules are most effective in the really bad times.
Here’s the drawdown chart that makes the point.
Aside from 2014, the timing strategy loses less than the Buy-and-Hold strategy, and often by quite a bit. And being able to avoid deep drawdowns is an investing superpower.
Here’s the performance summary of the Moving-Average Rule-based timing strategy versus Buy & Hold.
Even if you assume most or all of that excess return is lost to transaction costs and taxes, that you never had to deal with the pain of seeing your portfolio down more than 20% should be reason enough to consider a rule-based approach.
Ulcer Index is a numerical representation of the depth and duration of drawdown from previous highs. Higher the value, higher the level of “stress” (and hence Ulcers?) you would have endured. Read more about it on the creator, Peter Martin’s website.
And if you’re still reading, you should be asking: What is so special about a 100-day moving average? Why not 50. 150. 200. And do they even have to be multiples of 50?
What timing rule is right?
Simple Moving Average thresholds are just one way of setting up a timing strategy. Once you’re convinced of the need to have one, you could create your own composite metric. But we’re keeping things simple in this discussion and seeing what moving average threshold offers the most downside protection while not needing frequent exits and re-entries.
The chart shows results of setting varying moving average thresholds to time NIFTY entry and exit, from 10 to 250 days. y-axis is the annualized return. x-axis is the number of days of moving average used as exit and re-entry threshold.
The dark blue line is the Buy-and-Hold return that any “active” strategy needs to beat comfortably to be worthwhile.
Results suggest, the shorter the look-back period, the better the results barring some anomalies. If you only focus on annualized return.
A 10-day moving average threshold would return 13.9% annually versus 10.1% from a buy-and-hold strategy. But that would involve 293 trades which would significantly add transaction and tax impact. On the other hand, a moving average over 6 months seems to be counter-productive.
Don’t forget the impact of drawdowns. Chart shows the maximum drawdown when using the various moving average thresholds for entry and exit.
The green-shaded zone is the additional pain a Buy & Hold investor would endure compared to one applying a rule-based timing strategy. Every scenario results in a significantly lower drawdown compared to Buy & Hold.
Just that last chart alone, irrespective of the incremental promised return, should get every investor to question Buy and Hold.
Bottomline: Rule-Based > Discretionary
Deep drawdowns are gut-wrenching. They overload your rational decision-making capacity and make you second-guess every decision.
Add more of that stock that’s lost 40%? Or stay away from that falling knife? But then your returns from finding the bottom could be spectacular! Three C’s (Cash, Courage, Conviction and all that). It’s just very hard. Even if you know where the opportunities might be, the fear can be paralyzing.
One way to reduce the impact of market falls is through asset allocation. [Forget stocks for a bit, get your asset allocation right.]
Another is to tackle the aversion to taking losses and changing your mind. [Don’t be afraid of letting go. Changing your mind in a crisis panic fall]
But the most effective, and this is my opinion, is to use rules that you apply consistently.
The argument is not for using a 100-day or any other moving average threshold as an entry-exit signal but to have some signal that you can apply consistently to exit and enter a holding.
I had put forth a similar argument for individual stocks, albeit very mildly, in an older post: Does DMA-based timing score over buy-and-hold.
Our momentum strategy is an example of a completely rules-based strategy that decides whether and how much of a security to buy. But rules-based strategies do not necessarily mean high-turnover. The 100DMA NIFTY entry-exit strategy did 64 trades over 16 years. Hardly at risk of being termed a High-Frequency Trading Strategy.
In a separate post we will consider the question: What happens to our returns if we move into another asset instead of cash? Might a signal to exit an asset be the signal to enter another (Gold)? What might that do for our portfolio performance? And might that reduce our turnover?