This page has been archived and commenting is disabled.
Ugh! I Should Have Known!!!
This comes from the Department of I Should Have Known. More specifically, this is the kind of thing you do on a slow Tuesday afternoon before a Fed announcement.
Over the past several months, I have commented that the market will have an upward bias with a bid under the market and that shorting the market beyond an hour would be a very difficult proposition. Observing the intraday price action of the S&P Depository Receipts (symbol: SPY) over the past several months, it seems to me that markets never sell off for more than hour before the buyers appear. Not having much to do today, I decided to see if my observations were correct, so I crafted a simple study.
Study
1) 60 minute prices bars on the SPY
2) study start date: March 6, 2009
3) entry: two consecutive price bars where close less than open than buy next bar at open
4) exit: first price bar where the close greater than open and where close greater than entry price then sell next bar at open
5) no stop losses utilized
6) did not account for slippage or commissions
In the study, there were 67 trades yielding 30 SPY points. Buy and hold since March 6, 2009 has netted 38 SPY points. There was only 1 losing trade, but this is what we would expect from strategy that sold on the first close greater than the entry price. The average length of time of each trade was 8 price bars. With this strategy you spent 45% of the time in the market. The equity curve for this strategy is shown in figure 1, and this basically reflects the strong up move over the past 6 months.
Figure 1. Study/ equity curve
********
Figure 2. is the Maximum Adverse Excursion (MAE) graph for study #1. MAE measures individual trade drawdown (x axis), and what we can see is that 57 out of the 67 trades from this strategy had an MAE or draw down of less than 1.5%. These are all the trades (or carets) to the left of the blue vertical line.
Figure 2. Study/ MAE graph
*********
In other words, my observation that this market hasn't stayed down for long is pretty much correct. Over the past 6 months, two consecutive closes below their opens (on a 60 minute bar chart of the SPY) has generally brought in the buyers. Furthermore, with the average trade only lasting eight 60 minute price bars, traders didn't have to wait long before being made whole again.
- advertisements -
The facts, nothing but the facts. Great work.
What, no comments on this thread about the evils of short term trading? Sounds like a strategy that uses only technical analysis and likely would both buy and sell in the same day fairly often.
From comments I have seen on earlier posts on this site (which I dont agree with), this makes you a parasite - producing nothing for the economy. By trading short term in this zero sum game you are only taking money from long term investors a few cents at a time...how dare you!
Geez, with this cpu backtesting capability, you have an information edge, and seem to have a access to faster and more powerful information generated from computers than the average mom and pop at home. How dare you.
I have to call out this community on this...and say that you blur the lines on this stuff way too much. Flash and dark pools are bad - agreed. But, co-location is just one more step than the author of this post and his short term trading ideas. Further, the idea that HFT are parasites must be applied to this author - as he is producing nothing for the economy.
Food for thought.
Back testing always works great. The problem is only in doing the result in real time. The logical crutch you need to act will then proceed to beat your brains in and steal your money.
Good stuff. Clearly the actions of a million piranhas and the almost the complete absence of sharks or whales except as HFT's.
To me, the clearest point of deception is the lack of any real wealth being created.
While the ticker tells one story, of wild and unabashed rises in the most obscure and improbable of stocks, no where are there stories of new wealth, a day trader who has made $100 million with a $100 stake or the hedge fund on the verge of reporting a 3000% quarter. Certainly there were those who were rolled out as the villains back in October '08 but are there no hero's to be found this time around?
perhaps operators like that prefer to keep it that way.
Very nice work... interesting stuff... so are you still bullish on this market?
Thanks TTT. How many super computers did it take to execute this algo?
Thank you, sir. Good stuff. The trend is your friend.
nice. it won't end untilthis kind of stuff starts getting printed.
i've got one that needs to wait until an ex-div date passes in a few days
Good work... We'll prompt you to conduct this same study in reverse, when this market rolls over. Volatility is about to begin to expand. Stay focused.