Category Archives for "Research"
Continuing from the last post, I will show how using different definitions of passing our out-of-sample test can change our results. How luck can play a role if you use only one strategy to test in out-of-sample. How you split your in-sample(IS) and out-of-sample(OOS) can change results.
As long time readers of my blog know, I often use a market timing indicator in my strategies. My favorite one, and a simple one, is using the 200 day moving average on either the SPY or S&P 500 Index. I recently ran into these posts, Using Market Breadth To Gauge Market Health (Part 5) and Matt’s Breadth Indicator. Matt’s Breadth Indicator (MBI) intrigued me because I had not seen something like this and conceptually it is simple. I also liked that it was not “easy” to test or optimize on. Therefore hopefully not many people would be using this indicator and I could potentially find better values.
In Simple ConnorsRSI Strategy on S&P500 Stocks I showed a ConnorsRSI strategy on S&P500 stocks. In ConnorsRSI Strategy: Optimization Selection, I narrowed down the optimization to three potential variations that one could consider trading. This post will explore Sensitivity Analysis (also known as: Parameter Sensitivity) to help guide us on what to expect from each variation.
In the previous post, Simple ConnorsRSI Strategy on S&P500 Stocks, I showed a simple strategy which I optimized which gave 1,300 variations. Today, I will cover various methods to choose a strategy to potentially trade.
As anyone who pays attention to the market, the S&P500 is down nine days in a row. I had several people write me about this. I was talking to a trading friend over the weekend about this. Nine days down seems bad. Let us put this in a broader context. How far have we come down in those nine days? Only 3.07%. Now that got me thinking is 3.07% in nine days that bad?
My recent research has been on the volatility Exchange Traded Products. My focus has been on long trades using VXX and XIV. Although VXX has a very strong downtrend, I am not a fan of developing short strategies on it due to the huge upside risk. I wrote about XIV here and expressed some of the dangers of trading these ETFs.
I was at a recent talk of the Northwest Traders and Technical Analysts group where they presented a VXX strategy with some huge return and drawdown numbers. Trading this would be very difficult. This got me thinking. If I had a strategy like this, how could I tame the numbers? Through the years, I have seen various ideas about how to do this but never looked into it. Searching the web one can find various volatility ETF strategies with very high returns and high drawdowns. I found one that looked interesting and had lots of potential for optimization and improvement. Then, I optimized the hell out of it searching for a variation with over 100% CAGR. I found one but I would never trade it because I over fit the data. I needed to something to work with.
A reader sent this interesting link about Up/Down Capture. The great part about this article it is something I have intuitively known about my trading strategies but never tried to quantify. This was the bump I needed to investigate this concept. When the SPY moves up on average how much does my strategy profit? When the SPY moves down how much does my strategy lose? I had fun creating an interactive spreadsheet where you can change the numbers and see the results for your own strategy. This is one of the best spreadsheets I have done for the site.
Often one runs a optimization of a testing idea, then using some set metrics from these results, one picks a variation to trade. What often comes as a surprise to people, and myself the first time I saw this, is that your optimization runs are often mean reverting. What do I mean by this?
A common way to describe a mean reversion trade is a rubber band that stretches away and then snaps back. Something that Steve, my trading buddy, and I discuss when a trade keeps going against us is that the rubber band has broken. I have never tested that concept. Meaning after N day sell-off, are we now more likely to continue to sell off than bounce? Doing research is not always about trying to develop a new strategy but sometimes it is testing a concept. The concept may lead to a new trading idea.