jpb Veteran   Posts: 168 Joined: 5/11/2005 Location: Brown Deer, WI  User Profile | Here is what I'm in my final stages of testing: ---------------------------------------- Long condition: ((CASHFLW > 0) OR (CASHFLW > CASHFLW[LbPeriod])) AND ((EPS > 0) OR (EPS > EPS[LbPeriod])) AND ((OPINC > 0) OR (OPINC > OPINC[LbPeriod])) Short condition: C>5 ----------------------------------------- LbPeriod varies by market state but I began with it being 90. I began thinking about this when ATM using AutoTrade and a dynamic symbol list via OmniScan inserted SHOP (Shopify) into the list and traded it. SHOP had a number of days where it dropped. RTM strategies looked at this and said "Hey, look at that! A sharp down movement is likely to move back toward the mean. Let's get in!" The trouble with this is that SHOP is losing money and has for a long time. Why would any investor want to buy and hold this company? A trader might look at it and say, let's ride it back up, but an investor needs to feel the dream and believe the company is worth more. What I was really looking for was a list of companies that appeal to both traders and investors. In looking at the fundamental data OmniData has, I settled in on Cash Flow, EPS, and Operating Income. All three need to be either positive or better than they were in my lookback period. I began with a simple 90 day look back -- approximately a quarter but is really more like 4.5 months of daily trading bars. So, what results did I get? With no fundamental filter and data for the last 20 years: $10,000 grows to $1.77 million; MDD 17.6%; Avg ROI 33.4%; Avg MDD 8.5%; Calmar 1.9 With a 90-bar look back and 20 years of data: $10,000 grew to $1.34 million; MDD 13.7%; Avg ROI 32.1%; Avg MDD 6.5%; Calmar 2.34 With a lookback per market state and 20 years of data: $10,000 grew to $1.65 million; MDD 12.1%; Avg ROI 33.5%; Avg MDD 6.2%; Calmar 2.77 Changing the date range to be more current and less data shows similar patterns of improvement. Out of curiosity, I wondered what 1/1/2006 - 6/28/2010 would look like. I had a slight flip between my market state optimized filter and the static 90-bar for all market states, but they were within $250 out of $32k. For grins, here's a look at the equity curves. [Edited by jpb on 7/1/2019 2:54 PM]  Attached file : Screenshot_20190701_145039.png (42KB - 706 downloads) |
