Last night I revisited the topic of creating adaptive strategies for an audience from the AmiBroker Canada User Group and a local chapter of the Canadian Society of Technical Analysts. The presentation was well-received, and the audience asked thoughtful questions at the end. If you want to listen in, the recording can be found here.
Hi Matt,
1) How was the 35th percentile chosen for the Volatility criteria? have you tested other percentiles?
2) Doesn’t the HV100 ranking criteria bias the Volatile Bear and Volatile Bull regimes? Wouldn’t the use of ATR > $0.8 minimum for example give you range without volatility?
3) Other than this Market Regime and a Seasonal Market Regime, what others have you come across?
Many thanks for a great presentation!
LikeLike
Hi Ridwan,
For this presentation, my goal was to focus on the process of creating an adaptive strategy. Therefore, I did not spend time trying to optimize the definition of the different market regimes, but just used values that have worked well as filters in the past.
I’m not sure I understand your question about HV(100) biasing the Volatile regimes. I want to favor the stocks with the most volatility, as those frequently show the best mean reversion performance. We could use ATR instead of HV as a measure of volatility, but I would probably use something like ATR/Close rather than just ATR, because an ATR value of $0.8 is a lot different for a $10 stock than for a $100 stock.
There are many ways to define the market regime. For example, Dr. Van Tharp defines 6 different regimes: the four that I used in my example plus Neutral (sideways) Quiet and Neutral Volatile. One theme is that many regime definitions include a price component and a volatility component, but of course that doesn’t have to be the case. Even within those boundaries, you can use different lookback periods, different indicators and methodologies, different symbols to represent “the market”, HV vs. IV, etc.
Finally, it’s worth noting that you can apply adaptive strategy techniques based on something other than market regime. For example, you might change your strategy parameters based on the price range, volatility, or recent performance of the instrument being traded. While I can’t detail specifics, I’ve had some clients develop very successful strategies using this approach.
LikeLike