As noted in an earlier et post, that 'sophmore slump' is a common feature on enough tracking charts to be no surprise wherever they are. interesting to see the same feature on your chart too, nice 'junior jump' too by the way. Thanks again for all the good posts you do.
Fully Automated Stocks Trading
- Thread starter ValeryN
- Start date
- Status
As noted in an earlier et post, that 'sophmore slump' is a common feature on enough tracking charts to be no surprise wherever they are. interesting to see the same feature on your chart too, nice 'junior jump' too by the way. Thanks again for all the good posts you do.
Appreciate it! Hearing that it helps people is probably the biggest reason I keep posting.
I wouldn't make too much of it as I haven't discovered any systematic edge based on account equity or strategy performance itself. It always looks like there is something there though, I'd give you that
For instance my early performance this year on weekly and monthly basis looked too good to last and yet it continued where I expected it to end "for sure". We have an occasional exchange with Marsten on when we likely will see the beginning of a big DD or an equity peak. It a fun thing to do and let us be just humans, not emotionless machines, but I don't believe we ever stopped trading a strategy based on that stuff alone.
William Eckhardt gave a very good example of this. Not exact quote but in essence - cycle based methods work, if you let cycles extend/shrink or invert to make them fit any available data. Similar to that, "slump" can probably be seen anywhere, if you are dedicated enough to see it and flexible with time periods over which previous performance has occurred to justify its' significance.
+nothing goes up forever except JKS and my equity from Fed to March. Later one because it is a finite period.
Val
And I thought you don't trade penny stocks
Not sure if it can be considered a penny stock, but I had to check out of curiosity, and my own penny bot shorted it couple times, both profitably, though very short term like 24 hours:
Last edited:
Is your strategy that similar/related to Marsten's that you'd expect similar periods of drawdowns?
Some of our strategies are similar conceptually but different enough to lead to very different outcomes. Some are completely unique. The comparison I mentioned is not to see deviations, just a bit of social aspect, to share some pain and to share some joy. Marsten is the only trader I know who trades 100% in a mechanical way. That makes our problems much more relatable.
Glad to hear it worked for the bot!
It is a matter of penny stock definition. 20$ with mil $ daily volume is hardly a penny stock to me. Now it is 70$ with 10-20mil volume.
Val
My "penny" shorting bot doesn't actually trade only penny stocks as it trades the full universe of stocks. It just looks for best shorting opportunities, which happen mostly on sub $5 stocks, though it also shorts $500 stocks when opportunities arise. I call it penny stock bot because that's how it appears, and I'm even struggling with making it trade only higher quality stocks.
So now I'm actually curious how did you include JKS in your universe of stocks, or what do you trade? As I don't see it included even in Russell 3000?
Last edited:
The short strategy that picked up JKS trades full US universe provided there is enough liquidity and movement.
I specifically exclude what I consider penny stocks from that strategy as in real life they are often non tradable due to unable to borrow of short or lack of intraday liquidity = huge potential slippage.
While JKS was a loss the strategy itself made 26.8% YTD. It trades small enough that even big outliers are not that much of a big deal.
Attached equity curve for this strategy YTD at the end of this post.
PS. On a long side, in August, I launched a strategy that specifically targets penny stocks (again, what I consider a penny stock at least). While comparing my live vs model trades yesterday I discovered ~7k USD bug. Execution software wasn't watching all the stocks it was supposed to be watching, one exchange was inadvertently excluded. Otherwise it looks fairly consistent with my backtests. If not that bug, I could have bought a McDouble every day for 9.5 years.
Val
It took a bit to figure out what you meant by Just Killing Shorts
Agree 100% for systematic strategy. That would be too rare of occurrence also. Another reason - with such extreme outliers the probability of it continuing behaving in an unusual way also increases. It can be traded discretionary with a frequent reassessment of new data and acting accordingly with regularly updated probabilities.
I probably wouldn't have noticed it if mechanical system wouldn't have picked it up and took multiple losses on a way, which is fine, from discretionary perspective it just highlights unusual circumstances and an opportunity. Then I made a discretionary trade around 60$ expecting it to touch at least 20% off the highs within a few days. Drop didn't materialize and it started creeping up touching 70%, which was worst case scenario from my experience, it is the way it did that that made it such. Also, fundamentals for this market seemed like legitimately changed which was confirmed by weeks of persistent only green days. Even after most extreme volume days which normally would indicate a top all it could do is basically consolidate around highs. So I got out just under 70$ for 4 reasons
- At this point another extreme explosive move was likely
- Even the type of high probability correction I expected will now likely put me at a break even at best
- Based on historical precedents it could double from there in a climax move. Yes, it would have bigger correction in this case if happened fast, 30-50%, but I would be taking a lot of heat meanwhile and best case scenario would be losing 10-20% from the point of entry
- All of that would require constant monitoring of position and daily distractions
While looking at historical precedents of relentless upward action with ~10-30% per week for multiple weeks I found a true worst case which reminds that irrational can continue longer than one can remain solvent (if shorting in this case). Imagine how many people though that was a great short...
Get you popcorn ready, AAXN, 100x in 1y
Val
Not sure what was the best short ticker of the year by number of entries and win %, but I recently noticed Real Crappy Long. 5 short entries since April, 2 different strategies
That + Just Killing Shorts discussion highlights the nature of systematic trading - outliers will be on both side. Single trade, or even instrument, don't have much significance. 100x do, at some extent, better 1000s. Unless HFT. That's a whole different story. Single highly liquid instrument strategies also are, but I don't do that for many reasons.
Val
That + Just Killing Shorts discussion highlights the nature of systematic trading - outliers will be on both side. Single trade, or even instrument, don't have much significance. 100x do, at some extent, better 1000s. Unless HFT. That's a whole different story. Single highly liquid instrument strategies also are, but I don't do that for many reasons.
Val
- Status