Backtest: Reducing Momentum Drawdowns with Trend Overlay
This is a long post and may be truncated in the email app. For the best reader experience, it is recommended that you read this post on the “Substack” app or the website to zoom in on the charts and images.
Here are a few caveats before we move on to the backtest rules and results:
Only stocks listed on NSE’s “EQ” and “BE” segments as of 1st March 2024 are considered for backtesting
Data for delisted stocks is not considered in the backtests
Amibroker is used for backtesting
You might think that not considering delisted stocks would introduce survivorship bias in backtest results. However, the percentage of liquid delisted stocks is much lower when executing a backtest on all listed stocks. They won’t typically affect the backtest results significantly in the long term. The survivorship bias will affect the backtest results if done on a subset of the universe like the Nifty 100, Nifty 200, Nifty 500, etc.
Today, we are going to do a slightly different kind of backtest. The goal is to identify whether we can reduce drawdowns if our momentum portfolio has a trend-overlay cash call.
First, let me explain this a bit more in detail.
For a typical DIY momentum portfolio, we rebalance monthly or weekly and go into cash only if there aren’t enough stocks after applying filters. The only periods we went into cash in such a strategy were during the 2008 crisis and covid. The drawback of such a strategy is that we go into huge drawdowns as it takes time for cash to go out of the screens.
Last week, I wondered if we could go into cash with a trend overlay on our momentum portfolio’s NAV, reducing drawdowns while keeping the returns intact.
This post will try to answer that question and my findings.
Backtest Rules:
Rules:
Rank stocks by their average Sharpe returns of 12-9-6-3 months
Stocks should be above their 200-day moving average on rebalance day
If a stock is not above its 200-day moving average, then we exit that stock on rebalance day
The stock should be within 25% of its all-time high
Median daily volume in rupee terms over the past year should be at least one crore
1-year return should be at least 6.5%
The portfolio is equal-weighted
20 rupees per trade is considered for brokerage
Max stocks held is 30
The worst rank held is 60
As mentioned above, the rules are part of our typical momentum strategy. I performed a backtest for the above rules and then downloaded the equity curve of this strategy starting from January 2007 to March 2024.
We will consider the equity curve as the “NAV” of our DIY momentum strategy.
Next, we will run the backtest again, but this time, we will go 100% into cash if our NAV is below its 200-day moving average. We will re-enter if it goes above its 200-day moving average.
We will only check whether the NAV is below its 200-day moving average on our monthly rebalance day to decide whether to go 100% into cash.
Here are the backtest results of the same:
As the above results show, this reduced drawdown during the GFC (Global Financial Crisis) period. However, it underperformed in the post-GFC period.
Now, let us analyze the same but with different rebalancing periods and different parameters of trend-follow cash calls.
If you’re a free subscriber, you may upgrade and unlock the full post by clicking the link below.
Questions?
Reach out to me at 📩 viraj@momoindiascreener.in