Let's not let this subject die. Thanks for the input. If anybody does run any historicals please let us know. I use TradeStation, just have no idea how to accomplish it.
hmm... i have to say that trying to 'systemize' the end of the day anomalies/surprise moves etc...may be fruitless for all but the most advanced programmers. i think that discretionary trader's with many years of experience are best able to exploit these sometimes very profitable situations in the last 45-60 minutes of the day session. basically, they 'sense' the move, get on board when they have confirmation and RIDE it until the move TELLS them it is over, not where they THINK it should end. i was able to fortunately ride the wave up on friday afternoon from 86.25 on emini sp (i had a confirmed buy signal)--taking partials till my 92.50 area (fib extension/projection confluence) said to me the move was about done (under such heavy confluences in fibs--i had about 3 areas 'come' together) for my final 1/5th of my initial longs from 86.25. good trading to all... narc4ever
I do not agree. Many experienced traders can benefit from the end of day moves, not only the "most advanced programmers" or "discretionary trader's with many years of experience". All you need to do is research under which conditions these moves will occur and plan your actions accordingly. TM Trader
I don't know about predicting whether the move will occur or which way it will be, but its always been my understanding that one of the key reasons the end of day move occurs in the indices is due to mutual funds having to buy or sell equities depending on whether they have had a net inflow or outflow of cash for the day.
I've check this through several ST cycles... 1. When the market is in a bullish phase, the end-of-day run occurs almost every day. AND the sum of the runs is often greater than the entire cycle up-move!! 2. The end-of-day move can start ANY time in the last 2 hours (and sometimes a bit earlier). 3. When the end-of-day run is NOT occuring nearly every day, the market is either in consolidation or bearish. In fact, so long as the end-of-day run keeps showing up 80% of the time, it's a good reason to have a bullish trading bias. 4. The run on Friday might have been OE related.
Friday was a sideways day until the late breakout. The low was made pretty early, market looked weak, but by the time the late rally started there was already a succession of higher lows. Also look at how the bollinger bands compressed right before two upside breakouts. I'm looking at 3 minute candles, it's also evident on 5 minute. Today is different as we have much more follow thru on weakness. Also taking out minor low from Thursday, and SPY looking at the 50 day ma (daily). Definitely interesting.
Nice observations about bullish/bearish gnome and thanks for posting them. I'll have to check out what you say but rings a bell - we had EOD runs during the bull market but not on the bear market, and they're back in the current bull cycle. I always figured it was buy/sell programs hitting the market in the late afternoon when it has the greatest opportunity to move things into the overnight session, but while there is still sufficient liquidity. I recall that the estimate is that program trading makes up around 20% of daily volume. A lot of the moves start like clockwork at 3:45 EST. I picture some guy watching the clock, waiting to press the button on the program.