2

If I find that for stock X, for past 5 years of daily OHLC data, 80% of the time, the Open was at least 1% lower than the day's high.

If we find that in this 80% of the cases, the high was reached before the low.

So if I bought the open and set a limit sell order at 1% above the open. And a stop loss 2% below the open. If the stop doesn't get hit, then sell at the close.

So in 20% of the time, I lose a max of 2% and in 80% of the time, I make a profit of 1%. So is my expected profit at least (80 *1 - 40 *2) = 40%?

What is wrong with this strategy?

Improve this question
4
  • 1
    It is a great strategy. You are now classified as a day-trader. How much does each transaction cost you in commissions? (Day-traders might be charged commissions at different rates than the $7 ($9? $11?) per trade that so many firms advertise for more sedate investors.)
    – Dilip Sarwate
    Nov 27, 2014 at 20:38
  • Interactive brokers...1 $ per trade. And you can check your sarcasm at the door. Thanks
    – Victor123
    Nov 27, 2014 at 20:41
  • 4
    And if you had not checked your reading comprehension skills at the door when you visited the Interactive Brokers site, you would have noticed that (for transactions in the US) the fee actually is US$0.005 per share subject to a minimum fee of US$1 and a maximum of 0.5% of the trade value. (Fees for other countries and currency denominations are different).
    – Dilip Sarwate
    Nov 27, 2014 at 22:43
  • And don't forget the SEC fee on sales of US equities. That may or may not be included in your broker's fee schedule.
    – dg99
    Nov 28, 2014 at 20:57

2 Answers 2

Reset to default
7

Nothing is wrong and it should be profitable - but it sounds too good to be true. The devil is in the details and you have not described how you found those stocks.

For example, you may have scanned the 500 stocks in the S&P 500, and you may have found a few that exhibit that pattern over a given time window. But it doesn't mean that they will continue to do so. In other words they may just be random outliers. This is generically called overfitting.

A more robust test would be to use a period, say 2000-2005 to find those stocks and check over a future period, say 2006-2014 if the strategy you describe is profitable. My guess is that it won't.

Improve this answer
3
  • 4
    Also worth noting that a stop loss set 2% below the open doesn't guarantee you'll sell at that price, if it drops. A flash crash could see substantially higher losses.
    – ChrisInEdmonton
    Nov 27, 2014 at 23:54
  • 4
    Also, I'd imagine (but have not checked) that many stocks move up or down by a couple of percent in a day. A stock may close 1% up but may have gone 2% down during the course of a day. In that case, you lock in your 2% loss rather than your 1% gain.
    – ChrisInEdmonton
    Nov 27, 2014 at 23:56
  • 1
    "Past results are no guarantee of future performance." If you can't explain why your proposed strategy works, it probably doesn't and you're just seeing random ripples. (And, yeah, make sure you understand the fees before putting your money on the table.)
    – keshlam
    Nov 28, 2014 at 15:31
0

One should also point out that you make a major assumption in that the high of the day doesn't occur on a gap up in morning trading. It's unlikely that you'd fill at a reasonable price, thereby throwing your strategy into disarray.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .