Stocks & Commodities V. 31:12 (10-11): Letters To S&C by Technical Analysis Inc.

Letters To S&C by Technical Analysis Inc.

SELF-OPTIMIZAT ION OF INPUTS?

Editor,

I concur with your reader, Ronald Stuart, whose letter was published in the August 2013 issue. I, too, would like to see old strategies revisited (albeit with a modern computerized twist). I have been trading for 29 years, and have subscribed to S&C for two years now. I particularly enjoy the digital editions.

My current project is to place each strategy addressed in the Traders’ Tips section of your magazine each month on the same symbol and time frame (for example, symbol ANV placed on a 20-minute chart).

At the TradeStation forum, an inverse Fisher strategy is offered based on the May 2004 article in your magazine by John Ehlers (“The Inverse Fisher Transform”). I imported that strategy into TradeStation. I can even optimize the strategy inputs such as the buy line & sell line. I usually need to optimize the strategy inputs, as the backtesting and walk-forward trade analysis results did not appear very favorable. I open an additional workspace chart (same symbol, same time frame, which is a 20-minute chart) for each strategy from each month’s Traders’ Tips. I then optimize the inputs for each strategy.

Each strategy seems to have the same limitations: It will be profitable for a few days, then the inputs have to be reoptimized.

Then I read with great interest the article and related Traders’ Tips contributions for the self-adjusting RSI strategy as presented by David Sepiashvili in your February 2006 issue (“The Self- Adjusting RSI”).

I placed the self-adjusting RSI strategy on a 20-minute chart of ANV on July 24, 2013. It’s been running one week, and I can’t tell what I should be optimizing or even if I should be optimizing the length and/or K inputs. I see buy/sell short/cover on the chart.

What’s more, it seems to me that our computers should be able to optimize these inputs in real time on any given strategy. I read several technology magazines each month, and I am amazed at the amount of resources poured into predictive software and apps. For example, the Google Now app aims for “predictive intelligence,” offering assistance as users go about their day — from hailing a cab, to suggesting proper clothing (based on preferences and weather conditions), to local cuisine recommendations (based on previous consumptions), to timely reminders for personal grooming, and so on.

Real-time adaptive, automatic optimization of strategy inputs or self-adjusting strategy inputs could make great strides in placing the power of profitable trading in the hands of your readers.

P.S. What inputs should I be optimizing in the self-adjusting RSI strategy, or is it already set as a truly self-adjusting strategy?

Dave Sidwell

Thank you for your feedback. Regarding the self-adjusting RSI, it’s the overbought & oversold levels that are adjusted. The formulas provided in the article make these benchmarks self-adjust. —Editor