Thursday, February 25, 2010

Brainstorming Notes

Technical indicator constructed with the last X hours of data.

A lookback period of Y days.

Buy signals at level A in the indicator; sell signals at level B.

X and Y chosen to maximize profitability of the signals across a set of ETFs and A/B levels.

Trade the indicator for day Y+1 for the ETFs with the clearest signals, thus making the assumption that the next period will not differ significantly from the past Y periods.

Each day, reconstruct the indicator parameters over the moving lookback period and select a new set of ETFs to trade based on performance over the lookback period.

Repeat this process using daily periods for the indicator and weekly lookback periods. Using 5 minute periods for the indicator and hourly lookback periods. Etc.

If you do that, you find that Y is a small number, a smaller number than would allow for traditional statistical significance testing.

Which is what makes developing trading systems so difficult: before the market behaves in a pattern that can be detected conventionally, the pattern tends to change.

Which means that patterns have to be detected non-conventionally:

Predictability is a market variable that fades in and out over time.


The key is finding the time frame(s) where markets are displaying predictability/regularity.

Which would make a trader operate sometimes like a scalper, sometimes like a position trader, and sometimes like an investor.

Trading failure is a result of trying to make markets fit into a style of trading, rather than fitting the trading style to current market conditions.

8 comments:

Daniel said...

Three quick comments, Dr. Brett.

First, I agree with everything except the two following..

==> "thus making the assumption that the next period will not differ significantly from the past Y periods..."

--Change that to "NEVER making the assumption that, etc. etc."

THAT is the first filter. And there ARE metrics for such determinations-- your blogging colleagues (and links) Jeff P. and Rob H. for example use various objective measures of 'normalcy' and 'continuity' in such a manner..


The other one is a caution on the precise definition of a key term.

==> "..before the market behaves in a pattern that can be detected conventionally, the pattern tends to change."

Make sure you are quite careful in defining the term 'conventionally'.

Any system of detection MUST be strictly objective--conforming with its interior 'conventions'--or it isn't a system, but a set of general guidelines.

The detection 'apparatus' must NOT rely on intuition.

The trader applies his or her intuition TO the apparatus.

Good luck! You always make trading fun!

Daniel

Charles Upton said...

Very interesting post Doc.

The thought that comes to my mind after reading it is...

"Know when your edge is present and when it is not. When it IS there, PRESS THE HECK OUT OF IT!"

Of course deciding to not trade when it is not there requires patience, discipline, and also humility.


Charles

RDV said...

- I trade the ES, and the ES only.
- My timeframe is a combination of the 5- and the 15-minute charts.
- My most important indicators are the Nyse-Tick, Volume, the EMA20 and the keylevels from the daily and hourly charts.

Though I have enough flexibility to adjust my trading behavior to the regularly changing characteristics of the ES, I lack the equipment, the software and the brains (!) to trade in any other way and in any other products than the ES.

My trading succes lies in my patience to definitely NOT to trade in any other way or any other timeframe or any other product; the daily study of the markets and the necessary preparations; as well as the power to sit on my hands until I'm used to changed circumstances and/or characteristics.

That's the way a simple man as me looks at trading while making steady profits.

Trying to do it in any other way has cost me dearly. This with all due respect to you and others who do possess the brainpower to develop other ways of trading.

I suspect I'm not alone in this; I would love to see more reactions on this topic. Am I wrong? Should I do more?

Steveo said...

Great comment Doc,

We are in transition now, methinks. A few indicators back this up....non-conventional stuff.

My favorite indicators, DWC/GLD (Total Market / Gold Price), and EEM / SPY (Emerging Markets / S&P 500 ETF)

http://oahutrading.blogspot.com/2010/02/my-favorite-indicator-ratio-charts.html

Adam said...

Brett ~

Well done!

And true... time resolution vs. signal relevance... the holy grail of correlation... an elusive, shifty quarry.

Adam Sterling.

Adam said...

Daniel ~

"The detection 'apparatus' must NOT rely on intuition.

The trader applies his or her intuition TO the apparatus."

Dead right!

Adam Sterling.

Daniel said...

Adam- Thanks.

Steveo - Aloha, brudda.

Please help me annoy Dr. Brett-- who already is so busy he has to borrow sleep from others, since he can’t possibly have time to sleep himself-- to rework an intriguing short-term SPY/EEM reversion and trend indicator he has half done. It would dovetail beautifully with your longer time observation.

RDV - Yes you are very wrong to question yourself WHATSOEVER.

What Dr. Brett emphasizes ALWAYS is that the indicators are there to help us see more clearly. More and more of them is NOT nec'ly better; at some point more becomes distracting, rather than clarifying.

If you are truly gaining "steady profits" then you really ought do little or nothing other than keep it up!

That said, it was not clear whether your use of the "20EMA" was a 20ema of your 5minute or 15minute BARS,as an intraday trigger, or a 20ema of daily closes. Go back and read the post, you'll see it's ambiguous.

Good trading, all.

Daniel

Mark said...

RDV - I'm with you. I greatly admire traders who can juggle vast amounts of information at lightning speed like Dr.Brett. Unfortunately, I do not possess that kind of brain power or patience. Thus, I'm working on a simple "dumb guys" system using mean reversion. My hope is to become very good and very disciplined using this simple system to grind out consistent profits.