The Ultimate Strategy
The holy grail of returns would be if you could get equity like returns while having bond like volatility. While in theory such possibilities do exist, the question that arises is whether its feasible / practicable in execution where reality meets fantasy.
Motilal Oswal Mutual Fund has indeed launched one such fund aptly named – Motilal Oswal MOSt Focused Dynamic Equity Fund. In the words of Aashish Somaiyaa, CEO, Motilal Oswal AMC;
Now, as a systematic trader, my ears perk up when people talk about how great a system has been in back-test. Having build and buried thousands of systems that look great in back-test but fail miserably when it comes to the real world, there is a very high risk of the out performance being more of a mirage that seems to be out there but never can be reached.
To ascertain how long has the system been back-tested and how long its running in real time, I asked him details of the same and here is the reply;
The back-test period encompasses both bull markets and bear markets making at the very least it having some kind of validation about having shown its out performance when times were good as well as when times were bad.
But does that provide any sense of understanding given that any and every data mined strategy will pass the back-test with flying colors?
A better way to test would be to run the data through some statistical testing and see if it really is as good as it claims to be.
Unfortunately the Index is not really a Index as it is bounded and will move between a lower and a upper end making it more like a Stochastic.
But Motilal Oswal in a presentation of April 2016 has the following picture
While the comparison between Inception to April looks tremendously good, do note that much of the data is from the back-test. On the other hand, data from 2 Years on wards is of walk forward and hence more reflective of how good it is.
Below find a correlation chart of 100 period returns of MOVI vs future 100 day returns of Nifty 50.
Since the strategy is contra (Buy when market goes down, Sell when market goes up), do note that we need to focus on when strategy is positively correlated vs Nifty and when its Negative. On the face of it, it seems to do the job though the true picture can only be obtained after a few years of operation by this fund.
The key idea behind the launch seems to be with the idea that since Investors cannot digest large volatility (downside), by having a strategy that sells when markets are expensive and buys when markets are cheap, its volatility will be lower and hence hopefully the investor will stay longer.
In 2010, Motilal Oswal launched a Smart Beta ETF on the Nifty 50. The premise here as defined in the leaflet is as below;
What is MoSt50 basket?
MOSt 50 Basket is a fundamentally weighted basket based on S&P CNX Nifty Index (Nifty). The methodology is conceptualized and developed by Motilal Oswal AMC. MOSt 50 includes all the Nifty 50
stocks but not in the same proportion as Nifty. Weightage of a stock in MOSt 50 basket is determined by using Motilal Oswal AMC’s proprietary pre-defined methodology that assigns weights based on
stock’s fundamentals such as ROE, net worth, share price and retained earnings. This is to ensure that companies with good financial performance and reasonable valuation get higher weightage.
Once again, it was a good concept. After all, if you could buy more quality stocks and less of the bad quality, you theoretically should outperform the Index. The said leaflet also had a chart to showcase its advantage vs Nifty 50
As the chart and the accompanying table shows, the strategy was perfect. While you got the same volatility of the Index, returns were way higher. Since the above chart has both back-test data and walk forward, its tough to notice that it had started to deteriorate. So, how did it perform after being listed vs Nifty Total Returns and Nifty Bees;
As the chart clearly shows, the fund more or less under performed Nifty Bees most of the time and not surprisingly the fund house decided that it was going to follow the regular model from Oct 2014 instead of being fundamentally weighed.
This issue is not just of Indian funds. Take the ETF, First Trust Dorsey Wright Focus 5 ETF for instance. Using a concept of Relative Rotation among sectors that are showing strong momentum, the fund showcased how it has trumped S&P 500 returns
As the table showcases, it yielded strong +ve returns vs the S&P 500. Given that fact that in United States, majority of Mutual Funds under perform the Index, this was a real breakthrough or so the Investors must have felt.
So, once again, how has been the real time performance of the fund vs S&P 500
Once again, there is disappointment in store as forget out performance, for now the fund is not even performing in line with S&P 500. Then again, the time frame is short and who knows how it could perform in future.
Sometime back, NSE introduced a new index Nifty Quality 30. The fact sheet quotes “The ‘Quality’ investment strategy aims to cover companies which have durable business model resulting in sustained margins and returns”.
Once again, the idea is right. High Quality companies can and shall (in theory at least) out perform Low Quality companies and this is proved even in their back-test data.
Does the Nifty Quality 30 whip the asses of Nifty Total Returns. Holy Grail Unveiled, or is it?
Once again, its too short term to make conducive large term forecasts, but Quality 30 seems to be facing some strong headwinds as of now. Will it do better in future? I have no clue though if you believe, you now have the opportunity to invest in the Index through ETF’s such as Edelweiss ETF – Nifty Quality 30.
What is common in all of the three above examples is that the heart is at the right place – how to do better than Index with similar or lower risk. But the result is not as one would expect. Would Motilal Oswal fund be different?
Once again, I have no clue but would rather (as a Investor) wait for data rather than rosy forecasts / back-test data which may or may not be the best way to ascertain which strategy is good and which isn’t.
A known devil is better than an unknown angel.
Thank you for all the effort. You always seem to answer just the right questions that come to mind!
Very interesting analysis! So, what are the folks coming up with the strategies/funds missing – why do the back tests work for a long duration but start failing in real world? Are the back tests cooked up?
Generally its Data Mining. If you screen hundreds of factors and thousands of stocks, some will seem to be worthwhile but is just a random aggregation that may or may not continue in future