Scott Phillips: BST: 059: Reducing Drawdown - PursuitOfEdge/podcasts GitHub Wiki

  • uses daily bars
  • looks for objective things that you can't argue about e.g. an inside day
  • a breakout of an inside day in a sideways market tends to be fake, but in a trending market it's legit
  • hasn't found much better for a trend than just moving averages cross
  • a fakeout is a spike high (a bar with a lower high before it and lower high after it), then n bars which fail to break the old high, then a bar which breaks the old high, then sells short at the open of the range if it breaks the daily low with a stop at the high
  • retests which happen at changes of trends
  • gaps
  • in a trading range breakouts are probably gonna fail
  • in a trend pullbacks in an uptrend are an edge
  • or volatility edges when volatility reaches a historic low the next breakout is probably tradeable
  • zero edge in typical candle patterns
  • imagine a matrix:
  • bull | bear | sideways
  • high vol | low vol
  • the last 8 years (2008-2016) was a low vol bull market, which is best for just buy n hold, and countertrend setups have a dramatically lower odds of working
  • even perfect short setups just get screwed in a low vol bull market
  • low vol bear markets are the hardest markets to trade, the shortest to trade, and aren't really worth building systems for - last couple months before the bottom of 09
  • high vol bull market - 2001 blow off top, 2008 bubbling oil, crypto,
  • low vol bull markets are comfortable for pyramiding positing whereas a high vol bull market is usually the end of something preceding something else
  • sideways bull market 70s trend followers looked like geniuses due to stagflation and any inflation directed huge trends in commodities
  • low vol sideways market is tough to trade but you buy near the bottom of the range and get out quickly
  • vol: standard deviation which is just bollinger band width, ATR, average candle size, implied vol
  • sideways choppy looks like random walk, up up down, doji bars, represents bulls and bears are balanced and eventually one will punch out
  • sideways is below highs, upper bband is falling, etc.
  • he mostly trades trend and gets screwed in sideways markets
  • his systems are only 0.2, 0.11 (if he takes 100 trades he'll have 11%) expectancy which is very modest
  • he mentions sometimes his systems spend 10/12 months in drawdown. feels like he makes his money just 2-3 good runs a month and the rest of the time he's just losing
  • build systems with the market type as the base
  • trend: moving averages
  • sideways: lower highs, bollinger band is falling
  • some people do fancy stuff with equity curve like rolling expectancy/moving average, but it's usually not worth it and can create problems itself
  • likes shooting star pattern for breakdowns
  • bollinger/keltner breakout systems
  • 3-4 day time stop
  • Van Tharp: most traders put their stop too tight at the start of a trade, and too wide at the end of a trade
  • tighten stop dramatically after 4R
  • it is difficult to accept more than -30% drawdown, most people just don't have the balls for it
  • Dunn Capital best 40 year record, but it has had a -60% maxDD
  • everything that made him good as a businessman didn't help at all for trading (the idea that alphas can't trade)
  • being good at trading like being a good martial artist, concert pianist, chess grandmaster
  • book recs: Market wizards, rem of a stock op, van tharp, mark douglas, steenberger