Australian (ASX) Stock Market Forum

Tools being used by your trading competition

Jump to new
ThingyMajiggy said:
PM or email me if you want to keep the discussion going, especially if it's about me, people here wouldn't care for that.
Click to expand...


You may be surprised Sam .... you have been one of the most interesting people around ASF in recent years ...I would be very interested in your personal trading journey and trading strategies on many levels so please don't think the rest of the punters aren't interested in your story etc:cool:

Personally i don't fit your desire to see a long term successful trader, although I am now into my third successful positive year in succession but that is another story, and one I'm obviously happy about, but it is rightly tempered with great humility and respect of the markets! Will it continue? Who knows; I'm no Quant!!:)

I know you were doing some great stuff on the Dax, so I'd be the first to welcome another "Sam thread" on any or all of your musings in relation to anything trade related.

Cheers M8.
 
skyQuake said:
If a trader is truly successful, they run into liquidity issues. They can then choose to go wide eg. employ same strats using algos across more stocks/more markets (maybe even layer on diff strats) or go deep eg. investing. I think most people mellow out with age and kids etc so they go the investing route
Click to expand...

Yeah good point :xyxthumbs

barney said:
You may be surprised Sam .... you have been one of the most interesting people around ASF in recent years ...I would be very interested in your personal trading journey and trading strategies on many levels so please don't think the rest of the punters aren't interested in your story etc:cool:

Personally i don't fit your desire to see a long term successful trader, although I am now into my third successful positive year in succession but that is another story, and one I'm obviously happy about, but it is rightly tempered with great humility and respect of the markets! Will it continue? Who knows; I'm no Quant!!:)

I know you were doing some great stuff on the Dax, so I'd be the first to welcome another "Sam thread" on any or all of your musings in relation to anything trade related.

Cheers M8.
Click to expand...

Thanks barney! Well done on the 3rd positive year! :cool: Would definitely be keen to hear more!

Yeah I've still got some ideas to further build on the work I was doing in the Pull up ya DAX thread, whether it will come of anything or not is another matter, but interesting none the less :)

Thanks for the input!
 
Greetings --

A few of the comments posted cry out to me for response:

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

Supply/demand and sentiment drive prices, not computer programs.

We -- all people -- regularly try to assign causality. Correlation is fairly easy to detect, but causality is much more difficult. The best we often can do is react to the evidence. In this case, react to whatever prices appear in the history as we are developing trading systems and in the tick-by-tick transactions as we are trading. Computer program may or may not drive prices, but computer programs are the tools of choice for analysis.

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

If a trader is truly successful, they run into liquidity issues.

There are hundreds of equity related tradable issues that have daily liquidity on excess of 10 million dollars per day. SPY has liquidity in excess of 20 billion dollars per day. Many highly liquid issues have bid-ask spreads of one cent.

Following the advice of "the data prospector" outlined in both my "Quantitative Technical Analysis" book and "Foundations" book, select one tradable issue that has the characteristics that allow it to be traded profitably and work with it. To recap, those characteristics are:
1. enough volatility to offer some profit.
2. not so much volatility that risk is excessive.
Given that the issue passes both filters 1 and 2:
3. patterns in the data that precede profitable trading opportunities.

If liquidity is an issue, then either:
1. the trader is trying to trade issues that are intrinsically illiquid -- penny stocks with wide bid ask spreads, for example. Try sector-oriented exchange traded funds as an alternative -- XLB, XLE, XLF, ...
2. the trader is already a big player trading millions of dollars per day.

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

Variations of "some people make better judgements than do computer programs"

Both research and actual performance consistently show that well developed computer algorithms are much more accurate and consistent than the subjective decisions made by people.

Granted, some people have good results over some periods using subjective judgement -- and we hear about them. And many people do not -- and we seldom hear about them. The single book I recommend to everyone who asks "what should I read to learn about trading?" is Daniel Kahneman's "Thinking, Fast and Slow." Dr. Kahneman explains how, and in some cases why, we are so good at fooling ourselves -- the biases in our thinking that we ignore.

Those traders who are good at identifying global trends and selecting stocks that will benefit should continue to do that. Those who are successful due to luck in some form or another should plan for a time when performance suffers and decide how to recover. If the recovery will be an algorithm, the techniques I recommend will help. If the recovery will be subjective, I cannot offer helpful advice.

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

Python is just a language

Correct. It is the opportunities available for development and management of trading systems through use of Python that are exceptional.

For people who want to simply translate a decision tree trading model from TradeStation or some other development platform, I have published a line by line example. The result is the same trading system, with the added capability of follow-on analysis of risk and position sizing not available in a traditional platform.

For people who want to expand to any of twenty or more models, scikit-learn opens an extensive library of machine learning techniques. One that works well is an ensemble of thousands of individual decision trees combined as a "random forest" model.

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

Thanks for listening, Howard
 
OmegaTrader said:
Warren Buffet is a Quant.


You learn something new everyday.......
Click to expand...

Unlikely.

The geeky glamour of the quant took a big hit with the financial downturn, beginning with the so-called quant crisis in the summer of 2007, when quant funds took a nosedive. “All I can say is, beware of geeks bearing formulas,” Warren Buffett memorably told Charlie Rose in October 2008.

http://www.nytimes.com/2010/05/16/magazine/16FOB-OnLanguage-t.html

In my opinion, investment success
will not be produced by arcane formulae, computer programs or
signals flashed by the price behavior of stocks and markets.
Rather an investor will succeed by coupling good business
judgment with an ability to insulate his thoughts and behavior
from the super-contagious emotions that swirl about the
marketplace.

http://www.berkshirehathaway.com/letters/1987.html
 
Does Howard (or do you Howard) have some kind of track or performance record that is publicly available?
 
minwa said:
"...an ability to insulate his thoughts and behavior
from the super-contagious emotions that swirl about the
marketplace".
Click to expand...

That's successful trading right there. That one sentence. How to understand the potential for pain (losing) and pleasure (winning), and be relatively unaffected by the possibility of both... all the time retaining an intention to win.

It requires a complete re-wiring of the 'usual' mind. The usual mind is afraid of loss and desperately seeks to win.

Some people cope with that threat mathematically ("if I do x and y, then z is likely"), others by mental discipline.
 
howardbandy said:
Supply/demand and sentiment drive prices, not computer programs. (My post)

We -- all people -- regularly try to assign causality. Correlation is fairly easy to detect, but causality is much more difficult. The best we often can do is react to the evidence. In this case, react to whatever prices appear in the history as we are developing trading systems and in the tick-by-tick transactions as we are trading. Computer program may or may not drive prices, but computer programs are the tools of choice for analysis.
Click to expand...
Respectfully disagree. Another (direct) cause that drives price is news events. One less obviously causal is inside information which is not immediately obvious until the outsiders get the news.
 
Roller_1 said:
Does Howard (or do you Howard) have some kind of track or performance record that is publicly available?
Click to expand...

Interesting request. Howard is not here to sell you python courses....he's a recognized author and retired professor....why would he need to have a track record?
 
CanOz said:
Interesting request. Howard is not here to sell you python courses....he's a recognized author and retired professor....why would he need to have a track record?
Click to expand...

He means trading track record, not academic or career record. I'd like to know too. After all, Howard is right up there when it comes to systems trading, so it would give us all a good benchmark of what's possible. Even if Howard was using a bank of quantum computers to generate signals, it wouldn't mean anything without return$.
 
minwa said:
Unlikely.

The geeky glamour of the quant took a big hit with the financial downturn, beginning with the so-called quant crisis in the summer of 2007, when quant funds took a nosedive. “All I can say is, beware of geeks bearing formulas,” Warren Buffett memorably told Charlie Rose in October 2008.
Click to expand...

That was similiar to what Kerry Packer and Ashok Jacob told Mark Silbermann when he told them that his formula on Onetel's finances indicated that Onetel was headed for failure.
They then had Silbermann physically removed from the meeting because they thought they knew better.

Three weeks later Onetel went into receivership.

(Kerry also commented on ch9 that Merv Lincoln (StockDoctor) was just another idiot and didn't know what he was talking about when he reported that by his StockDoctor formula Onetel was doomed to failure)
 
An excerpt from an interesting article posted on November 3rd 2016. History never repeats?

"Someone Is Wrong" - Why The Quant War Means Vol Is Set To Soar

An example of just this "war of the robot (traders)", is that as Credit Suisse observes, a "quant war" is quietly taking place, and as Bloomberg eloquently adds, "computer is fighting computer in a hedge fund tilt that could get messy regardless of the winner."

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

The problem facing the quant world is learning how to respond to the same signal inputs. For the most part, the responses have not led to a wide divergence, however something appears to have changed this time around: "as the quant population exploded, analysts have spent more time trying to pin down how they will react to volatility and other inputs. And with good reason, as long-short funds are sitting on $215 billion in assets, while CTAs oversee some $330 billion, according to BarclayHedge, a database that tracks hedge fund performance."

The major divergence observed above is troubling to Connors, because it is reminiscent of the periods before the two most recent corrections, when funds took similar opposing bets. Trend-following commodity trading advisers and equity long/short funds are currently 47 percentage points apart in their U.S. equity exposure, the largest spread since just after the S&P 500 Index bottomed in February. Before that, the last time the spread grew to more than 40 percentage points was in August and September of 2015, the worst two-month stretch for the S&P 500 since 2011.

The causes for the variance in quant sentiment have been isolated: CTAs bearishness is being spurred by an event occurring after the election: the Federal Reserve’s December meeting, when rates may rise, Connors said. On the other side, equity long-short managers have been buying neglected companies like banks in anticipation of higher rates, he said due to relative valuation mispricing.
Click to expand...
http://www.zerohedge.com/news/2016-11-03/someone-wrong-why-quant-war-means-vol-set-soar
 
The Fundamental champion is Buffet
The Quant champions Massive hedge funds

But it's about being able to use fundamentals and increasingly data analysis in YOUR trading
That doesn't remotely resemble the above.

I just can't see how people can't grasp this opportunity.

Oh well.
 
tech/a said:
The Fundamental champion is Buffet
The Quant champions Massive hedge funds

But it's about being able to use fundamentals and increasingly data analysis in YOUR trading
That doesn't remotely resemble the above.

I just can't see how people can't grasp this opportunity.

Oh well.
Click to expand...

I think some are missing the point, that it's not about algos/HFT vs discretionary, but using data analysis in/with your own trading. I guess a lot of it is statistics work and helping find/validate an edge or just improving your chances of finding profitable trades. From what I've seen of machine learning, guys are just using it in a directional sense, if it can give you a strong hint about whether tomororw, or next week will be up or down, that's a fairly big hint that you could potentially position for.

I'd like to see this stuff in action, but to do that I'd have to learn/do it all myself no doubt, which I'm in the process of, seems to be lots of discussion on it but not much evidence...yet. I'm all ears/eyes, if there's something that can improve your trading, I'm not sure why you'd want to shut it down just because it's an algo/ML and you're adamant about the algo vs discretionary war.
 
Good grief....there are so many different time frames and markets that are traded...seems a broad brush to tar everyone with the same feathers here.

Regarding intra-day trading. You can clearly see algo activity at certain times of the day/session. Anyone worth thier salt has always known that there are times of day to trade, Algoes still need baby sitters. If you trade intra-day know when to trade, applies to discretionary and algorithmic-ally....
 
As long as the futures market has an order book with 100% transactions hitting it, i think the opportunities will be there for alert, trained eyes....if transactions somehow migrate away from the book, to some kind of 'dark pools' all bets are off for the little guy.:2twocents
 
howardbandy said:
What would substance be?
Click to expand...
howardbandy said:
As I write in my "Foundations" book, the days of chart reading, long term holding, and simple trading algorithms are over. The business of trading is changing with astonishing speed. It is now about applied mathematics, machine learning, Bayesian statistics. Traders without skills in math, programming, statistical analysis, and scientifically developed trading techniques are at a severe disadvantage.

Best regards, Howard
Click to expand...
Hello Howard. Substance? Evidence in the form of results or more to the point, evidence in the form of trading algorithms or methods.
 

Similar threads

howardbandy
Getting Started in Machine Learning for Trading
2 3
Replies
50
Views
21K
entropy
E
jjbinks
Python - getting started
2 3
Replies
43
Views
19K
fiftyeight
fiftyeight
darkhorse70
The Beginner Trader's Manual Guide
Replies
9
Views
20K
commoyo
commoyo
Top