Any Math Whiz Here??

[IAS_LLC]

Thanks but I kinda knew that already.

Changing the parameter wasn't what I was talking about. I was thinking more of an "adaptive" moving average based on different conditions.

You describe a "recursive Bayesian estimator". Kalman filter for the linear-gaussian models....particle filters for non linear non gaussian.

 

[d08]

Yes it is possible (I'm a math whiz).....Back testing is overrated.

 

[themickey]

......but there is a better way - read my thread 'trading is easy'... instead of price action you do situation analysis, narrative analysis... that's how my boys play the crowd... now your success rate gets a boost... ignoring the narrative is like pretending you are trading in a vacuum and just staring at the screen with the price candle going up and down... you are leaving a useful on the table - the narrative.

and this method will not be crowded, because YOU are simply better than the rest... every story every narrative is never the same but you see thru the truth never the less.

Doz, can you give a brief synopsis on what you mean by
situation analysis, narrative analysis.
Regarding your thread 'Trading is easy', hmmmm, I don't wish to have to read the whole thread to just get to this point.

 

[Real Money]

Real Money said: Back testing is overrated.

Hi d08. Yea you read that right. I think backtesting is overrated. In fact, most people are not even remotely qualified to perform rigorous scientific analysis of historical data. This is quant stuff, for quants.

I know there's a lot of successful traders that rely on backtesting, and all other kinds of trading systems analysis. These people are likely to fairly rate the utility of such analyses.

Things that are underrated:

Lead/Lag analysis
Arbitrage Monitoring
Proprietary Trading Techniques
Market Structure

and any other thing that may yield a structural edge. Backtesting should verify an edge. But finding an edge is harder than selecting a strategy that backtests well. That's all I mean here. It's my opinion, and I am referring to discretionary trading more so than automated systems.

 

[vanzandt]

The above chart shows that the crossover for the EMA line took place 2 bars prior to the SMA line. This got me thinking whether it's possible to hack the formula to make the lines respond faster, eg. faster crossover.

Try darts.

 

[d08]

Hi d08. Yea you read that right. I think backtesting is overrated. In fact, most people are not even remotely qualified to perform rigorous scientific analysis of historical data. This is quant stuff, for quants.

I know there's a lot of successful traders that rely on backtesting, and all other kinds of trading systems analysis. These people are likely to fairly rate the utility of such analyses.

Things that are underrated:

Lead/Lag analysis
Arbitrage Monitoring
Proprietary Trading Techniques
Market Structure

and any other thing that may yield a structural edge. Backtesting should verify an edge. But finding an edge is harder than selecting a strategy that backtests well. That's all I mean here. It's my opinion, and I am referring to discretionary trading more so than automated systems.

That reads more like "most people don't know how to accurately backtest". I agree that it's far from just choosing a strategy that backtests well (although supposedly this is what RenTech does).
But I don't sit around for months either to imagine some grand strategy just to find out in a backtest it does not work at all. I have a rough idea what to look into and even if my initial idea was completely wrong and the opposite is true, I still follow through. I've discovered some interesting things just by looking at data relationships without any initial idea for a strategy.

 

[ZTrader888]

The above chart shows that the crossover for the EMA line took place 2 bars prior to the SMA line. This got me thinking whether it's possible to hack the formula to make the lines respond faster, eg. faster crossover.

The answer is no. The reason is that an indicator prints when it receives the price data. It is not a fortune teller. When you say "faster", you can set your indicator to "on bar close" FALSE. However, this will trigger the indicator to print on all incoming data and not on the close of the bar. This will generate a lot of false signals.

You said that the EMA precedes the SMA by 2 bars. You're misinterpreting how the indicator works. The EMA gives higher weighting to recent data while the SMA weights all data.
I'm not going to write out the math but I strongly suggest that you read this link to understand what moving averages do and don't do.
https://en.wikipedia.org/wiki/Moving_average

My suggestion is to write a code for your crossover and then test it on different size charts. You're only using averages which makes it simple. I have to also say that it's naive to think that combination of moving averages is going to crush the market. It doesn't.

 

[SunTrader]

Using MA's, to trade price itself, no matter how complicated/convoluted the formula is like trying to catch smoke.

 

[.sigma]

ema sma... all is garbage... sooner you stop sooner you get to profitability, if ever..

I agree assuming one is using these lines to actually enter/exit trades.

But we all know price reacts to the 200/50/22 day moving averages regardless of past movement.

 

[SunTrader]

...
But we all know price reacts to the 200/50/22 day moving averages regardless of past movement.

Sometimes - but when it does it is coincidental and self fulfilling based on how many are sold on the idea they have any value.

A line is just a line no matter the math involved and represent very little to do with true support and resistance based on prior price and volume levels.