can price action predict market moves

longandshort said:
Are you billy hwang?!

Lol but seriously, you making money on your trading has nothing to do with you using a bad signal. You are making an attribution error by explaining your returns with your chart signal. What are you are likely profiting from is liquidity provisioning or intra-day momentum/reversal. For the former, the best signals are liquidity pre-trade, and for the latter, it's the slope of returns.

I've sat on a trading floor on both the sell side and buy side for 8+yrs, I've seen the gamut of strategies and no one uses charts to trade because they are very bad at representing what is actually going on. They are, however, visually appealing.

Good analogy would be a chef applying salt to their meal. "Grab and sprinkle salt until you start to see them pooling" is inferior to saying "eyeball 1 tspn" which is to using a teaspoon to measure and apply your salt.
More...

I can make a list of price levels as to where to buy and where to sell. They will be valid for months to come. They will work. And then you will claim random walk. Let me say that for the AMD trade, the price already reversed very close to my level before going back. Under your theory, the move already occurred so it was only random that it worked this time. So why does it keep working? The drawdowns should also be relatively random but that isn't the case.

Multiple times I've caught the GME low of day. The one time it failed resulted in price getting barely back to breakeven the next day which is the maximum length of time allowed to let a trade allow for the Event B/breakeven effect. This so called randomness appears a lot less random when its the HoD/LoD which is statistically significant portion of the time.
 
1. How often do prices hit those levels you predict
2. How often do you trade on the signal
3. If you remove the return from liquidity provisioning and stock trend, how much of the return is attributed to your signal?
 
longandshort said:
1. How often do prices hit those levels you predict
2. How often do you trade on the signal
3. If you remove the return from liquidity provisioning and stock trend, how much of the return is attributed to your signal?
More...
1. Perhaps 6-7 times per week, roughly
2. More than 80% of the time. The times I do not enter, it is usually because I already had entered on a different, yet correlated signal (eg I'm short AMD so I won't take a short SPY signal). Other times I just miss the trade.
3. What timeframe do you define as trend? I can be up a lot in a bull market on pure short trades only.
 
so what you're telling me is that you are using charts to make predictions on prices and that trading is working out for you. if you can't illustrate your returns ex-post or provide a study of your signal (going back in time, probably at least 100 min examples, likely more) then you are not describing an actual strategy but some gut feel you have based upon your interpretation of the chart. there is no way of proving the validity of such a strategy, because it exists a priori lol and is not empirical. that it works for you is not proof of its validity.
 
longandshort said:
....and no one uses charts to trade because they are very bad at representing what is actually going on. They are, however, visually appealing......
More...
upload_2021-4-1_9-32-44.png

Al Brooks is a perfect example, he looks like a Pom but with that arrow on his nose he's most likely Red Indian.
 
longandshort said:
so what you're telling me is that you are using charts to make predictions on prices and that trading is working out for you. if you can't illustrate your returns ex-post or provide a study of your signal (going back in time, probably at least 100 min examples, likely more) then you are not describing an actual strategy but some gut feel you have based upon your interpretation of the chart. there is no way of proving the validity of such a strategy, because it exists a priori lol and is not empirical. that it works for you is not proof of its validity.
More...

screen-shot-2021-03-31-at-6-56-23-pm-png.256013

Here is one proof, but not huge sample size
https://www.elitetrader.com/et/threads/prediction-based-trading.351176/page-42 Here is the other proof. All entries and exits called in real time and timestamped with no edits.
https://www.elitetrader.com/et/attachments/screenshot_20210221-085540-png.252717/ 1 month account statement I posted in another thread a while back
 

Attachments

  • Screen Shot 2021-03-31 at 6.56.23 PM.png
    Screen Shot 2021-03-31 at 6.56.23 PM.png
    270.6 KB · Views: 133
  • acctvalstat.png
    acctvalstat.png
    42.2 KB · Views: 15
longandshort said:
this is not proof of your signal. great returns and good job otherwise!
More...

Cheers. But, the signals were the only ones profitable; when I tried to high volume scalp on C2, the executions weren't what I expected and ended up in losses; should have been up 2X% on the month at least.

How would one go about proving the signal without spilling the sauce?
 
You must log in or register to reply here.
Back
Top