Investment by getting exposure to risk factors goes by a few names, most notably factor investing and/or smart beta. I explore using proprietary trading methodologies to create an active portfolio strategy that achieves high exposure to momentum, and its back-test reveals significant outperformance of both S&P500 and the most liquid momentum ETF (MTUM). Throughout this article, I will also sprinkle bits of information about the inner workings (but no secret sauce) of proprietary trading.
This active portfolio strategy is inspired by a proprietary trader who is trading a variation of this strategy in Asian equities. As a prop-inspired strategy, this active strategy exhibits very high risk (volatility of returns) and high draw-downs. Proprietary trading firms manage this risk by using a "diversification across strategies" to keep risk manageable. Individual investors should go in with the awareness of high risk, and should not put all your eggs in one basket.
Smart Beta and Momentum
Smart Beta emphasizes capturing investment factors and rotating the portfolio holdings to stocks with a higher amount of these factors. Common smart beta factors are size, value, quality, liquidity, dividend yield, and momentum, among other factors. I shall focus on building a portfolio with exposure to momentum in this article.
Momentum has a few definitions, but the definition and implementation that works for us here is that stocks that have high returns in one time period tend to have high return in the next period. This simple definition saves us the trouble of having to do complicated calculations regarding returns of multiple time periods and/or adjusting those returns by risk/volatility.
(An interesting point to note is that proprietary trading strategies tend to be based on simple ideas or simple price patterns like breakouts or gaps. The difference tends to be how an individual proprietary trader defines it, for example Trader A might define a price gap to be more than x% from last close, while Trader B might define it as X times of ATR.)
Investment Universe and stock selection
Our investment universe will be the 505 stocks in the S&P500. The stocks will be grouped into 11 sectors by the stock’s GICS sector. I will also simplify the analysis of this investment portfolio by assuming the current constituents of the S&P500 have been in the index since the start of the data period, and that stocks that do not have price data in earlier time periods will not be in our consideration set. (While this is not ideal, accounting for the S&P500 constituent changes will increase our data set by more than 40% to 700+ stocks, and complicate the stock selection process during each rebalance.)
Stock selection follows the very simple rule: at the end of a period, look at the performance of the stocks in a sector, choose the best performer that has the highest return among all stocks in the sector, buy and hold this stock for the next period.
To illustrate, the following table shows the 28 stocks in the S&P500 utilities sector (the smallest sector for illustration purposes) and uses adjusted close prices from Bloomberg. Bloomberg's adjusted close prices are adjusted for dividends and corporate actions, which causes the many decimal places, but the result is that your holding period return calculation will include capital gains and dividend returns, instead of being just capital gain returns.
(Stock list from Wikipedia: List of S&P 500 companies, prices from Bloomberg)
The result is that NRG shows the highest return for the year of 2017 at 113.69%, and thus we shall invest in NRG for the next time period, the whole year of 2018. At the end of 2018, the final result is that we realize a 39.59% return from our investment in NRG. For reference, below is the price chart for NRG over the 2017-2019 periods. Again notice the difference between Bloomberg adjusted prices above (adjusted for dividends) and the actual traded prices below from Investing.com.
From here we will repeat the process and select the stock with the highest return for the year of 2018, to invest in for the year of 2019. In this case, it just happens that among all the stocks in the utilities sector, NRG again gave the highest return for the year of 2018, giving us a nice realized return for the year, and instructing us to invest in NRG for the year of 2019.
With 11 sectors, and applying the same methodology to select 1 stock from each sector, we end up with a portfolio of only 11 stocks. At any point of time, we will have only 11 stocks, 1 from each sector. We will also allocate an equal weight to each of the 11 positions of our portfolio.
(In a proprietary trading setup, this is ideal because we know exactly how many positions we have at all times, and the number of positions is constant making for easy position and trade monitoring. As for the equal weight, it makes the P&L/return calculation easy because the portfolio return is the combination of the individual trade returns, and the individual trade returns can just be combined using arithmetic average.)
In running the back-test of this strategy, I picked an arbitrary start date (01 Jan 2001) and use data up till Jun 2019 (end of the most recent quarter). While 2001 and 2002 were down years for the S&P500 and the economy was in a recession, this is not the reason why the start date was chosen (it was arbitrary). If a strategy has an edge over passive buy-and-hold, then time-specific conditions in the market should not matter, and the strategy should outperform its benchmark index. That being said, negative returns are reasonable expectations for a long-only strategy in a down market.
The compounded annual growth rate (OTCPK:CAGR) of this strategy is almost 18%, with an initial investment of $1 turning into $16.58 by the end of 2018. Risk-adjusted-return represented by the Sharpe ratio is a respectable 0.63. However, I’d like to point out the very high risk as represented by a 49.86% drawdown in 2008 during the financial crisis. Imagine losing half your portfolio value in a year! In comparison, the S&P500 CAGR during the same time period is 4.7%, dropped 38.5% in 2008, and had a Sharpe ratio of 0.27. (If you see different CAGR numbers elsewhere, it is probably due to the different starting time period of the data. Remember I used an arbitrary start date of Jan 2001, which happened to be the bursting of the tech bubble.)
Now the Explosive Momentum Portfolio outperformed the S&P500 in both returns and Sharpe ratio, but it is also useful to compare it to the most liquid momentum ETF MTUM.
MTUM has only been in existence for 6.5 years, and has not been through the terrible period of 2008. Comparing the Explosive Momentum Portfolio to MTUM in the time horizons up to 10 years, the Explosive Momentum Portfolio showed a higher return, and a marginally higher Sharpe ratio for all time horizons. This is honestly surprising to me, since MTUM uses a more sophisticated method to calculate momentum taking into account the volatility of returns, and that I could not find any information on how often they rebalance their holdings.
So a common question asked in proprietary trading is "if what you are doing has an edge, can you do it more?" Let's further analyze the explosive momentum portfolio strategy by increasing the rebalance frequency to quarterly.
(Asking the above question is realizing that proprietary trading is a numbers game that obeys statistics. When you have an edge in the market, the outcome of a trade may not always be positive/profitable, but it follows a distribution that has a positive mean. By the Law of large numbers, the more you do it, the closer the average of the results is to the expected value, or the distribution mean, which is positive.)
The CAGR of the quarterly rebalancing explosive momentum portfolio is a cool 26.39%, with an initial investment of $1 turning into $71.82 by June 2019. The Sharpe ratio is also a respectable 1.03. The S&P500 and MTUM are presented below. (The below statistics include the first 2 quarters of 2019, which were not included in the yearly data above, hence the slight difference in annualized values. If I ended the price data series at 2018, the annualized statistics for the quarterly data would be exactly the same as the yearly data.)
While the explosive momentum portfolio outperformed the S&P500 in both return and Sharpe ratio, it outperformed the MTUM only in return (26.39% vs. 15.94%). MTUM had a higher Sharpe ratio for the longer term horizons. This is not surprising as MTUM had a good 2019, and had minimal losses in the 2015-2016 period.
What surprised me was that by increasing the rebalancing frequency, the explosive momentum portfolio strategy increased its CAGR significantly from 17.96% to 26.39%. I shall look at a few specific cases where certain stocks gave huge returns or losses within the timeframe of being invested in them.
AMD in 2Q 2018 increased 49.15%, and then surged 106% in 3Q 2018. This is due to optimism in its business model related to gaming and cryptocurrency mining, and also after a few analysts issued favorable research reports on AMD.
(Chart from www.investing.com)
NFLX in 4Q 2012 increased 70%, and surged 104% in 1Q 2013. This is due to earnings showing video-streaming subscription growth.
(Chart from www.investing.com)
(Chart from www.investing.com)
NKTR in 1Q 2018 surged 78% due to a lucrative collaboration deal, but fell 54% in 2Q 2018 after pessimistic medical data of its products. So momentum does not always carry a stock, and negative returns still exist within the distribution of outcomes (win rate is not 100%).
(Chart from www.investing.com)
An example of a typical trade without any fundamental news (that I know of) moving it would be like COG, 8.14% rise in 2Q 2017, and a modest 6.88% rise in 3Q 2017.
(Chart from www.investing.com)
Generalizations of momentum
While the majority of trades act like the COG example above, there are a good proportion of cases where the stock would rise significantly in one quarter, giving us a signal to invest in them, and then "exploding" upwards in the next quarter (giving this strategy its name of explosive momentum). While examining every single case is too time consuming, I shall attempt my own interpretation of these cases and why quarterly rebalancing appears to be the best time horizon for rebalancing.
I hypothesize the following: most stocks in the S&P500 are constantly followed by knowledgeable investors (both Wall Street and main street). In the quarter where they perform the best in their sector (and giving us the signal to invest for the next quarter), these knowledgeable investors combine their industry knowledge with new developments/news about the company that was just released, and formulate their view that this stock would make a good investment, and they start buying. Sometime later (maybe the next quarter), more news gets released about the company, confirming the views of these knowledgeable investors. At this point, other investors who might not have the deep industry knowledge, or did not initially share the views of the knowledgeable investors, start buying into this view, and start buying the company's stock. This creates big buying pressure pushing the stock price up. These other investors probably finish their buying within the quarter, allowing us to take profits and look for the next trading opportunity. It is possible that the average rate of such market-moving news is quarterly.
The above is just a hypothesis, and I appreciate any inputs, ideas, or suggestions that any reader can contribute.
In this article, I created a momentum-based portfolio strategy, with many characteristics of proprietary trading strategies, including simple systematic rules for buying and selling, application of the same rules to many sub-group of stocks (GICS sectors in this case), constant number of positions, equal weights, and attempts to increase the frequency of trading. The best performing version of the explosive momentum portfolio strategy is the quarterly rebalancing version, with a long-term portfolio CAGR of 26.39%.
While the return is high, this strategy has very high risk, characterized by the high volatility of returns and high draw-downs (-49.86% in yearly rebalancing, and -37.03% in quarterly rebalancing). Hence, I do not suggest using this strategy for your entire portfolio, investors allocating a small part of their portfolio to using this strategy should be well aware of the risk. Investors should not just take the strategy as-is from this site, but make any adjustments for your own unique situations, transaction costs, or risk appetites. And of course, do your own due diligence regarding stocks and strategies before risking any money.
(A long-term CAGR of 26.39% might appear high to buy-and-hold investors, but is actually pretty low for proprietary trading strategies. Proprietary traders who trade leveraged instruments like futures, on top of using massive leverage on the capital, can hit annualized returns in the hundreds of percents. Also, individual strategies might have high risk, but proprietary trading firms have a portfolio of many uncorrelated strategies. Think of diversification across less-than-perfect-correlation stocks, but in this case the diversification is across uncorrelated strategies.)
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: As a prop-inspired strategy, this active strategy exhibits very high risk (volatility of returns) and high draw-downs. Individual investors should go in with the awareness of high risk, and should not put all your eggs in one basket. Investors should not just take the strategy as-is from this site, but make any adjustments for your own unique situations, transaction costs, or risk appetites. And of course, do your own due diligence regarding stocks and strategies before risking any money.