No and I don't have to work at this point. I love the intellectual challenge it provides and long term compounding effect on wealth. It's counter intuitive but I do believe I'm a better trader because I do something else full time.
Fully Automated Stocks Trading
- Thread starter ValeryN
- Start date
- Status
No and I don't have to work at this point. I love the intellectual challenge it provides and long term compounding effect on wealth. It's counter intuitive but I do believe I'm a better trader because I do something else full time.
I can agree that it makes you a better trader at this specific style of trading. Hundreds of other trading styles could distract your brain and provide similar benefits. I'm also guessing you might have other interests.
Anyway good to see that a Canadian brother is performing so well.
Discretionary overrides is not something I typically do these days, but on occasion, when odds seem to be overwhelming and I happened to look at the market it can happen.
Typically when volatility is high = more fear/irrational behaviour. Then a record is made in my log and $ outcome is tracked to keep myself honest about value over time. Last year all of them were good calls. Could be luck of course. If not, I think part of it due to lack of pressure to make them. My living doesn't not depend on it, I trust the systems too, but they can't handle every single edge case perfectly. Like driving using GPS typically works, but if a driver can clearly see a roadblock ahead might as well take an earlier turn.
That is of course, if the driver has eyes and is not delusional.
Not sure what's analogy for systematic trader. Perhaps making sure that similar probabilistic apparatus is used as for system creation. No emotions or greed involved, just numbers crunching to size up a very particular situation in front of ones' eyes, which system might treat in a too general sense otherwise.
Not gonna go too much into details but this year 3 overrides were done. To give readers a sense of what would trigger
My personal notes are more detailed and with $ PL outcomes, just trying to share general thinking process around those overrides.
Val
Typically when volatility is high = more fear/irrational behaviour. Then a record is made in my log and $ outcome is tracked to keep myself honest about value over time. Last year all of them were good calls. Could be luck of course. If not, I think part of it due to lack of pressure to make them. My living doesn't not depend on it, I trust the systems too, but they can't handle every single edge case perfectly. Like driving using GPS typically works, but if a driver can clearly see a roadblock ahead might as well take an earlier turn.
That is of course, if the driver has eyes and is not delusional.
Not sure what's analogy for systematic trader. Perhaps making sure that similar probabilistic apparatus is used as for system creation. No emotions or greed involved, just numbers crunching to size up a very particular situation in front of ones' eyes, which system might treat in a too general sense otherwise.
Not gonna go too much into details but this year 3 overrides were done. To give readers a sense of what would trigger
- On Jan 4 I felt my short exposure was a bit too low for the market we were in and a slight adjustment was made to one system to be more short on average. Without position size increase.
- On Jan 21st I felt market went too fast too much and overrode one minor rule that kept more longs over the weekend than it normally would. I crunched some numbers and probabilities of market ending up much lower seemed way less than around the same or higher on Monday.
- On Jan 28th' I closed two short positions manually and did same override for longs as in #2. There is a giant high volume reversal on weekly and some extremely oversold stocks can take off for a short term trend, and those 2 happened to have lost ~40% already in less than a week. Longs are not likely to suffer severely over the weekend. Another consideration was - long/short exposure seemed way off. I was looking at 6 short vs 0 longs coming into the weekend. Had 4/2 after adjustments. So a bit of manual balancing around edge case market scenario.
My personal notes are more detailed and with $ PL outcomes, just trying to share general thinking process around those overrides.
Val
Was up +5.5% couple of days ago for Jan. Vacationing since, didn't look at the market or systems. Noteworthy - all PL came from longs.
Bandy's books in this order:
- Mean reversion
- Modelling trading systems performance
- Quantitative trading systems
Take strategies from the mean reversion and code them, in RealTest or AB. Aim to have at least 500 trades per test. Anything less is likely non representative, the more the better. If you see something viable try executing manually 100% like a system would and see if you can, if costs match your assumptions. Just use small $ account to minimize cost of mistakes and if all is working as expected gradually increase size / allocation. You might be able to have something running in a couple of months this way.
Over the time you'll add more testing methods, strategies and tools. No need to wait till you master all. It will never be complete.
Don't ramp up your live trading size too fast after you start. $ fluctuations will mess with you. They will make you quit or change your system at the worst possible time. Takes time to get used to them. Always longer than you think. Ideally never think in $ terms but %.
Good luck.
Val
This one post is like a roadmap for a beginner. Thank you for your generosity and time, especially during your vacation. I am sure I'll come back with more questions. Kudos to your system's success on the YTD performance.
Thanks.
Despite rather big down day it seems like stocks are far from a real sell-off. Was coming into today with almost no long exposure and 10 short positions, not even a single one had moved down much. I don't recall when, with such a high short exposure, literally nothing made any profits to speak of. To me, that indicates there is still not much fear out there still.
Meanwhile weekly SPY seems to perfectly positioned for further downside move.
Portfolio is 5/0 short/long currently. Happens to reflect my discretionary view but it's not intentional. Systems make their decisions in accordance to their pre-programmed logic.
Meanwhile weekly SPY seems to perfectly positioned for further downside move.
Portfolio is 5/0 short/long currently. Happens to reflect my discretionary view but it's not intentional. Systems make their decisions in accordance to their pre-programmed logic.
Last edited:
While many NDX stock being destroyed I looked at a pretty vanilla NDX rotational system for any resemblances with the previous tech bubble.
Current period we're in - 2020-2022 Feb looks very similar to 2000-2002 Feb. I can backtest this system over 30 years and for any periods other than those two it didn't not hit more than 8% DD. During those periods that happened 2 times within a span of a year, so DD pattern becomes somewhat of an identifiable characteristic. Last time a prolonged period of inactivity followed and it took it nearly 5 years to reach new highs accompanied by a general market decline for 2.5 years.
I run that system on a small account and expect it to be out of market for awhile soon. Might do a discretionary SQQQ long there.
Current period we're in - 2020-2022 Feb looks very similar to 2000-2002 Feb. I can backtest this system over 30 years and for any periods other than those two it didn't not hit more than 8% DD. During those periods that happened 2 times within a span of a year, so DD pattern becomes somewhat of an identifiable characteristic. Last time a prolonged period of inactivity followed and it took it nearly 5 years to reach new highs accompanied by a general market decline for 2.5 years.
I run that system on a small account and expect it to be out of market for awhile soon. Might do a discretionary SQQQ long there.
Ib does get stingy with margin
- Status