fotchstecker
304 posts
msg #155661
- Ignore fotchstecker
modified |
2/9/2021 1:38:51 PM
(...aside from real backtesting that is...)
I think there's a real problem with stockfetcher filter performance values (the Perf(%) column in the filter results tables).
The "performance" column is misleading. It uses the close price from signal day (or evening) as the basis of the performance percent-change going forward.
The problem is that if you judge the value of your filters based on eyeballing the PERF% column, you are using a datapoint that doesn't translate to the real world. Close data (and settled OHLC daily data) is after the fact. Prices are finalized after market close. You can't enter until the open the next day, which is a different datapoint. OPEN prices are almost invariably different from prior CLOSE prices, and this delta is hugely significant.
(You could do things like run filters intraday and try to buy at the end of the session, but you would be using unsettled data and likely incorrect symbols to buy. I've done this. It wasn't accurate at all because final prices changed the final filter results.)
------>You can't enter at a close price and therefore you can't expect to get the same price dynamics as the Perf(%) column is indicating. This is a BIG DEAL.
STOCKFETCHER!
What I'd like to see is a second default performance column based on the OPEN of the following day -- the time when you'd actually place real-world trades. I realize that this will require an extra day of data to generate, but it's MUCH CLOSER to what actual performance is like.
It can't be hard to add this and it is a much less misleading depiction of a given filter.
|
xarlor
587 posts
msg #155688
- Ignore xarlor |
2/10/2021 8:32:59 PM
I get what you're saying and it has merit. However, I have used many strategies that require an entry at the close (most Larry Connor's for example). I simply put in an Extended-hours limit buy at the closing price of the day. They are almost always filled. This doesn't work for OTCBB tickers, but I stay away from them anyway.
|
fotchstecker
304 posts
msg #155700
- Ignore fotchstecker
modified |
2/11/2021 8:58:36 AM
Thanks for commenting, Xarlor. I agree that the strategy itself dictates whether MOC, after-hours limits, or on next session make sense.
Have you found that after-hours orders are generally advantageous? I've wondered but never studied it. I did read a finance journal study once about the significance of overnight price change, so I might test it out.
I use a custom tool for orchestrating live orders based on filter CVSs. This exists because I need to assemble baskets and daily manual entry of these would take a long time. I can't recall but I think there might be a limit order capability in it. It was a long time since I developed and can't remember what I did with this last version. I mentioned this because I'm thinking aloud: I might be able to enter after-hours limit orders with it in the same way...
|
xarlor
587 posts
msg #155709
- Ignore xarlor |
2/11/2021 8:10:03 PM
Never, ever put market orders for extended hours, you'll lose your shirt. You can probably get better pricing after-hours if you don't mind the real possibility of not getting filled.
My mentality on these trades is "I want to enter a position at the close of the day." If after-hours trading is actually trading lower than that price, I honestly don't care, because in a perfect world, I would have entered at the close price regardless.
That said, there are a few times that TDA gives me a much better fill than the limit buy I entered, but those are exceptions. I've had very good luck getting filled at the exact close of the day when putting that as my limit buy during extended hours. Sometimes it takes until 7:30pm ET before I get filled, but they normally do fill.
|
fotchstecker
304 posts
msg #155714
- Ignore fotchstecker |
2/11/2021 11:26:33 PM
at my broker it's not possible to enter a market order in extended session. limit only.
|