Tim Sykes and Penny Stocks

[guru]

That is what I'd have expected rather than the equity crash described in the original post.

To be fair, most of penny stock strategies do lose money when automated, due to inability to process press releases, no level 2 data, larger time intervals while some traders make decisions in seconds, bad quality historical pricing data, incorrect stock split and dividend data, and most importantly: large short squeezes that can blow up any shorting strategy.
So I also run the risk of blowing up, though I mitigate it by small trade sizes, no different than trading options where you wouldn’t short naked with more than very small % of capital.

 

[Gaslight Capital]

To be fair, most of penny stock strategies do lose money when automated,

Indeed, for all the reasons you give and more. No where have I ever seen Tim Sykes claim that his methods and strategies are meant to be automated.

 

[caacapital]

I think It’s a valid strategy for the most part Any stock that gets on penny stockers watch list are the stocks are trading Tens of millions of shares On the day even. However these stocks get these moves they are moving at that point it’s about getting in and getting out just like any other stock option or futures contract. Should anyone pay for these strategy’s when they are given out for free? Well that’s up to the individual whatever strategy will take time to master.

 

[zenostiffler]

Haven’t looked at yours yet, but I’ve tested over 50 million penny stock strategies, every variant imaginable.
Tim Sykes switches his trading styles from shorting (that’s where he made most money initially) to going long, like in the current environment. Though still shorting opportunistically.
Are you mostly shorting or going long, or both, or switching between them?

I was merely testing what Sykes claims to be trading, both long and short. Turned out his "systems" don't work. Assuming he described them fully in his various uninformative emails. I rather doubt Sykes has 50m systems and I do not believe he does any back testing. Although clearly he should. His stuff seems to be based on ultra simple technical analysis applied in a rather unsystematic manner. But I have never worked with the fellow. Perhaps I am doing him a dis-service?

 

[zenostiffler]

Indeed, for all the reasons you give and more. No where have I ever seen Tim Sykes claim that his methods and strategies are meant to be automated.

All systems which use "rules" can be automated. It is a pity that Sykes has not tried describing his rules to a computer. There again perhaps he has and he has some hidden genius he has not disclosed to his followers?

 

[zenostiffler]

I'm not going to adapt your code because anything done with QuantConnect is crap, but if you want to be proven wrong, here are my live stats of a penny stock strategy, with all trades streamed live (free):
https://deustrader.com/plutus_shortus.html
View attachment 230054

This is a sample older demo strategy that's been running since the start of the year. Internally I trade different variants that I can modify and debug on the spot. Though for now I'm more busy with a new system for trading options.

I find Quantconnect to be an excellent piece of software, but each to his own I guess. What exactly is Deustrader? The details on your website ares somewhat scant. Is it a signal service you are selling (or intend to sell)?

 

[zenostiffler]

Did you take into account him being paid in stock on the backend? Read the fine print.

No - but then is he actually being paid as a promoter? Perhaps he is? If so, he does not disclose that to his clients.

 

[zenostiffler]

I think It’s a valid strategy for the most part Any stock that gets on penny stockers watch list are the stocks are trading Tens of millions of shares On the day even. However these stocks get these moves they are moving at that point it’s about getting in and getting out just like any other stock option or futures contract. Should anyone pay for these strategy’s when they are given out for free? Well that’s up to the individual whatever strategy will take time to master.

If so, then Sykes hides his methods quite well. Applying his disclosed rules on a systematic basis seems to provoke disaster. But I could easily have misinterpreted the information given out or indeed miscoded it.

 

[guru]

All systems which use "rules" can be automated. It is a pity that Sykes has not tried describing his rules to a computer. There again perhaps he has and he has some hidden genius he has not disclosed to his followers?

One of his initial "rules" years ago, was to spend hours if not days analyzing press releases and financials, to look for signs of fluff/vaporware/sensationalism in press releases, and signs of invalid business model. Do you have such "rules"?
One of his current rules, I suspect, may be to distinguish companies strongly affected by coronavirus, either positively or negatively, or sensationally (thermal cameras, supposed cures, etc). Do you have such "rules"?

Some of this could be done subconsciously, he may recognize one stock vs another, similarly looking press releases, etc.

His other basic published rules based on indicators, indeed can do very little.
And, I do believe that Sykes has lost his edge, because even his own systems and rules work during some market conditions but not others. I have some strategies that are similar to his that would've worked amazingly well during some years, but would've lost everything in different years.
He may be able to recognize some of those market conditions, and increase his marketing efforts when he's able to capitalize on those conditions. He may sit more quietly when he isn't doing well.
Recently penny stocks are doing amazing well for going long, so I also see increased activity from some traders making tons of money. This may all quiet down when market conditions change, I mean when their victims (newbie traders) run out of stimulus money

 

[zenostiffler]

To be fair, most of penny stock strategies do lose money when automated, due to inability to process press releases, no level 2 data, larger time intervals while some traders make decisions in seconds, bad quality historical pricing data, incorrect stock split and dividend data, and most importantly: large short squeezes that can blow up any shorting strategy.
So I also run the risk of blowing up, though I mitigate it by small trade sizes, no different than trading options where you wouldn’t short naked with more than very small % of capital.

Quantconnect does not have level 2 data but is does have tick data and bid/offer data. Grattini admits he makes little use of any clues offered by the depth of the book - I wonder in any event whether spoofing goes on here which could render the book misleading.

My suspicion is that the granularity offered by Quantconnect would at least give some hint at profitability or otherwise.

As for automation, let's not forget Simons. Admittedly we can't test at this latency but I would have thought that any reasonable level of granularity would be able to give an indication as to whether what Sykes is doing is profitable or not.

As for press releases and SEC filings, these can of course be read into an algo these days. I did not do that but it may prove an interesting exercise.

Perhaps there is a more discretionary element to Sykes which makes all the difference?