Building a trading bot?

[FrankieK]

Pretty difficult to make any money there. I would say that most that claim to be profitable were lucky enough to ride out a long term trend for a while. There's a lot of survivorship bias in the industry.

That there is 4 shore. Which is why I personally do not touch individual stocks. I stick to dividend paying funds, Closed Ended Funds to be precise, and the VIX. A bit odd-ball but it works.

 

[FrankieK]

Apologies Havoc for the thread creep.

 

[HavocXphere]

Starting to think my time is better spent analysing shares direct than spending time coding something...pity.

Apologies Havoc for the thread creep.

No worries. Happy with any kind of chatter in the decidedly quite biz subforum.

That there is 4 shore. Which is why I personally do not touch individual stocks. I stick to dividend paying funds, Closed Ended Funds to be precise, and the VIX. A bit odd-ball but it works.

How do you filter out the closed ended ones?

 

[Thor]

Starting to think my time is better spent analysing shares direct than spending time coding something...pity.

+1

 

[FrankieK]

Starting to think my time is better spent analysing shares direct than spending time coding something...pity.


No worries. Happy with any kind of chatter in the decidedly quite biz subforum.


How do you filter out the closed ended ones?

Pretty much like one would with ETFs. Seeing that Closed Ended Funds (CEF) is the older brother of ETFs the process is roughly the same. Diversify and pick sectors where you think there will be growth. Drawback is the costs, usually run from 1.5 - 2.5%. Upside is the dividend, which can be as high as 18 - 20%.

A good website to investigate is www.cefconnect.com and of course investopedia.

 

[HavocXphere]

Pretty much like one would with ETFs. Seeing that Closed Ended Funds (CEF) is the older brother of ETFs the process is roughly the same. Diversify and pick sectors where you think there will be growth. Drawback is the costs, usually run from 1.5 - 2.5%. Upside is the dividend, which can be as high as 18 - 20%.

A good website to investigate is www.cefconnect.com and of course investopedia.

Familiar with the concept - just didn't realise there is a retail market for them.

Regardless - interesting but not gonna work for me.

 

[FrankieK]

Familiar with the concept - just didn't realise there is a retail market for them.

Regardless - interesting but not gonna work for me.

Ja, CEFs is not very "hip". They are fairly liquid cause they trade like a stock on the NYSE but there is a lot of investigation to be done with them. They are definitely a niche product.

I think a good product for a trading bot would be the VIX, for the simple reason that the thing always mean reverts. If it spikes then short it, if it is very low then buy it. But, IMO, a good understanding of Implied Volatility on options is also required.

 

[Beachless]

Starting to think my time is better spent analysing shares direct than spending time coding something...pity.

You will need to do that at the end of the day anyways. If you code first you will have the framework but not the knowledge if you research first you will still need to put the framework together. Either way this is a long term project with ever changing criteriea. Not really a holiday project but there is an increadible amount you can learn.

 

[bchip]

Starting to think my time is better spent analysing shares direct than spending time coding something...pity.

Your time will be best spent on finding a method to tackle the markets that work with your personality,
ie if you want to be a famous musician and you dont like singing then dont waste your time going to choir practise,
try out the drums or the flute instead, despite the advise given on this thread.

Find a methodology that works for you as I can show you succesful traders with most methodologies,
there are those who trade charts that do well, there are those that code that do well, and guys
who read balance sheets that do well.

Focus on the one that interests you and keep at it for about 5 years.
Its 5 years if no-one helps, its faster if you get mentored by the right person and slower by ...others.

Only you will be able to find which "instrument" works for you, as it depends on your background
and your personality...for example IT guys like coding, most people like charts because you can read it
in seconds and some people even like reading balance sheets (go figure!)

Just understand there is no right and wrong way with your method (Im obviously using that loosly)
but the crux of it comes down to how much you make when your right and how much you loose
when your wrong.
Van K Tharpe proved that you didnt need to use any funny formulaes or statistics or indicators
as he outperformed the market 3 years in a row with his dartboard and coin toss experiment,
where he use a coin to determine whether he enteres a trade.

 

[cguy]

Your time will be best spent on finding a method to tackle the markets that work with your personality,
ie if you want to be a famous musician and you dont like singing then dont waste your time going to choir practise,
try out the drums or the flute instead, despite the advise given on this thread.

Find a methodology that works for you as I can show you succesful traders with most methodologies,
there are those who trade charts that do well, there are those that code that do well, and guys
who read balance sheets that do well.

Focus on the one that interests you and keep at it for about 5 years.
Its 5 years if no-one helps, its faster if you get mentored by the right person and slower by ...others.

Only you will be able to find which "instrument" works for you, as it depends on your background
and your personality...for example IT guys like coding, most people like charts because you can read it
in seconds and some people even like reading balance sheets (go figure!)

This is terrible advice. There are definitively trading methodologies and styles that are becoming progressively harder to make money from, and ones that are emerging that are easier to make money with. That some people make money doing something doesn't have any bearing on whether or not one should do it or not (it may be near impossible to do, or soon will be near impossible to do). Similarly, what interests you could be a dead end road too. Rather try to understand the landscape, and where things are going - don't just do something because it sounds interesting and other people are doing it. Get a lot of different perspectives, and make sure that they're from viable sources.

Just understand there is no right and wrong way with your method (Im obviously using that loosly)
but the crux of it comes down to how much you make when your right and how much you loose
when your wrong.
Van K Tharpe proved that you didnt need to use any funny formulaes or statistics or indicators
as he outperformed the market 3 years in a row with his dartboard and coin toss experiment,
where he use a coin to determine whether he enteres a trade.

The only things this proves, is that there are people out there with a very poor understanding of statistics and markets. There is so much wrong with the above statement, it beggars belief.

 

[bchip]

This is terrible advice. There are definitively trading methodologies and styles that are becoming progressively harder to make money from, and ones that are emerging that are easier to make money with. That some people make money doing something doesn't have any bearing on whether or not one should do it or not (it may be near impossible to do, or soon will be near impossible to do). Similarly, what interests you could be a dead end road too. Rather try to understand the landscape, and where things are going - don't just do something because it sounds interesting and other people are doing it. Get a lot of different perspectives, and make sure that they're from viable sources.



The only things this proves, is that there are people out there with a very poor understanding of statistics and markets. There is so much wrong with the above statement, it beggars belief.

I think you may need to start substantiating your statements here rather than just vaguties.

For example you said that nobody can trade charts and yet Garth Mackenzie and Peter Brandt have been
outperforming the market year on year using charts and they are competlely transparent with their trades.

Van K Tharpe proved 3 years in a row that a coin toss experiment outperformed the markets
because his hypothesis was that it comes down to how much you risk vs how much you make when you win.
Which is the same thing that George Soros said.

Finding your personality on the market is also advocated by Van K Tharpe, Market Wizards - Jack Swagger
and Mark Douglas, trading in the zone.
All these people are huge successes in the market and the advise is consistent,
and then theres ...your advise on the other end.

Taking a step back, just to get this right if you take a cricket team and you take a guy like Alan Donald
who was a brilliant bowler but the worst batsmen in history, your advise is that he must
simply practise batting (and forget bowling completely) because you heard that Jacques Kallis was a great batsmen?

 

[cguy]

I think you may need to start substantiating your statements here rather than just vaguties.

For example you said that nobody can trade charts and yet Garth Mackenzie and Peter Brandt have been
outperforming the market year on year using charts and they are competlely transparent with their trades.

Firstly, I never said any such thing.

You should ask yourself how unusual is it really to outperform the market? You state this like it is some sort of proof that they possess some sort of deeper strategy/ability. Mindlessly investing widely enough allows you to break even with the market. Investing selectively, but naively will give you an expectation equivalent to the market drift, but the variance will mean that periodically one is likely ahead. Trading at a low enough frequency, these periods can last years, and can just as easily be better or worse than the market. Focusing on a particular time interval, can make even these naive approaches look lucrative, when it's actually really just bad stats.

I know it's possible to beat the market, but are you sure that those you think are beating it due to charting are not just a result of selection bias or survivorship bias? Also, are you sure that the way they're beating it is the way you think they're beating it?

Van K Tharpe proved 3 years in a row that a coin toss experiment outperformed the markets
because his hypothesis was that it comes down to how much you risk vs how much you make when you win.
Which is the same thing that George Soros said.

I don't think you understand what the coin toss experiment actually did (it was about position entry only IIRC). I also don't think you realize how unremarkable it is that this happens. It's also, really, really not so much a "proof", rather than "something that would have happened if things were set up just right". Do you remember the Bible Code?

Also, the rewards vs risk metric isn't either of their hypotheses, it's a well known idea that has been around formally since at least 1952.

Finding your personality on the market is also advocated by Van K Tharpe, Market Wizards - Jack Swagger
and Mark Douglas, trading in the zone.
All these people are huge successes in the market and the advise is consistent,
and then theres ...your advise on the other end.

I expect that the personality your're likely to find from those 1989-2000 era books, is likely to be a naive one. Times have changed radically since these were published.

Taking a step back, just to get this right if you take a cricket team and you take a guy like Alan Donald
who was a brilliant bowler but the worst batsmen in history, your advise is that he must
simply practise batting (and forget bowling completely) because you heard that Jacques Kallis was a great batsmen?

This analogy comes from a gross misunderstanding of what I'm saying, because for some reason you think I'm offering advice to some set of elite chart traders, rather than someone with near zero trading experience who is trying to figure out the path that will most likely offer them success.

 

[reactor_sa]

I met a guy from Deutsche Bank back in 2007, we were on some development training if I recall, anyway he had developed a trading bot that picked up really small trades but in volume. Made the bank a few million every month with minimal risk. It's definitely been done before.

 

[cguy]

I met a guy from Deutsche Bank back in 2007, we were on some development training if I recall, anyway he had developed a trading bot that picked up really small trades but in volume. Made the bank a few million every month with minimal risk. It's definitely been done before.

Yes, but with millions of dollars of infrastructure, and teams of quants and a team of world class programmers. Also, it was much easier in 2007.

 

[bchip]

@cguy
Im trying my best to understand your point of view
I think clearly theres some miscommunications here

Some of the things you mentioned to me now, were the things I mentioned to you...earlier

Regarding the charting comment

Firstly, I never said any such thing.

I got the impression from here

Or they trade they the chart.

Pretty difficult to make any money there.
I would say that most that claim to be profitable were lucky enough to ride out a
long term trend for a while. There's a lot of survivorship bias in the industry.

For sure. My point is that a lot of those who think they are the
former are actually the latter, they just don't know it.

I know it's possible to beat the market, but are you sure that those you think are
beating it due to charting are not just a result of selection bias or survivorship bias?

This was coincedentally from Garths email this morning
"Last year the show managed a profit performance of 25% ...
That comes on the back of returns of 43%, 42%, 52% and 53% in the prior 4 years."

At what point does this stop being luck and start being skill?
Maybe Garth understands something that most dont.

--------------------------

I agree that there are "turns" on methodologies, that at certain times
some of them work better than others (like value investors suffering in SA past 8 years),
I understand that but I also dont understand how you justify its "terrible advise"
telling someone to become good with a strategy that works with their personality.

don't just do something because it sounds interesting and other people are doing it.

...?
You understand that thats exactly what Im saying...someone should find their own way
with a methodology that works with their personality.
(which you said is terrible advise)

Im really trying my best here, I think my question is more...what point are you trying to make?
Because your comments are allover the place now, and some are running in circles.

Is it that reading information from news is a highly probably way to success?

what method do you think the average person is currently doing...? Its exactly that.

 

[cguy]

@cguy
Im trying my best to understand your point of view
I think clearly theres some miscommunications here
Some of the things you mentioned to me now, were the things I mentioned to you...earlier
Regarding the charting comment
I got the impression from here

There is a very definite qualitative difference between the complement of "most"/"a lot" and "nobody". The one explicitly claims possibility, and the other one explicitly states impossibility (which isn't even knowable, so this most certainly is not my stance).

This was coincedentally from Garths email this morning
"Last year the show managed a profit performance of 25% ...
That comes on the back of returns of 43%, 42%, 52% and 53% in the prior 4 years."

At what point does this stop being luck and start being skill?
Maybe Garth understands something that most dont.

It requires a deeper statistical analysis of the specifics of the decisions and markets to determine the likelihood luck or skill. I have no idea what he is or is not capable of without the previously mentioned analysis. Five years of good returns is far from sufficient evidence though.

I agree that there are "turns" on methodologies, that at certain times
some of them work better than others (like value investors suffering in SA past 8 years),
I understand that but I also dont understand how you justify its "terrible advise"
telling someone to become good with a strategy that works with their personality.
...?
You understand that thats exactly what Im saying...someone should find their own way
with a methodology that works with their personality.
(which you said is terrible advise)

You explicitly said "Focus on the one that interests you", and gave examples of how people can be successful in any of a few mentioned trading styles, and suggest that matching your personality with a trading methodology is what matters. I actually disagree with all of this - it's the equivalent of "follow your passions".

I believe in finding out what makes the most sense to do, and seeing if one can cultivate an interest and ability for it. There are some methodologies that are becoming prohibitively difficult to trade due to competition, market conditions, technology, cost/skill barrier to entry, etc. it makes no sense to run after these because of interest or "personality" match.

Im really trying my best here, I think my question is more...what point are you trying to make?
Because your comments are allover the place now, and some are running in circles.

- Home algo trading is very very difficult, and has become progressively harder for a variety of reasons, and will continue to do so
- The same for chart trading
- People with poor statistics backgrounds see skill where there is none
- Statements like:

"Van K Tharpe proved that you didnt need to use any funny formulaes or statistics or indicators as he outperformed the market 3 years in a row with his dartboard and coin toss experiment, where he use a coin to determine whether he enteres a trade."

are dangerously naive
- A deep understanding of a sector is the best way to invest for someone without professional investment experience who is willing to spend time on beating the market. This goes way beyond reading news, and looking at company financials.

Is it that reading information from news is a highly probably way to success?

No.

what method do you think the average person is currently doing...? Its exactly that.

... and the average person is almost certainly following the market drift as a result.