Anyone managed to build a scalping/HFT-strat?

We all know that the magic of compounding and uncorrelated portfolio trading is the “holy grail” to get straight R² and quickly amassing profits.

Trading frequency matters. Anyone succeeded in the glorious attempt to produce robust scalpers?

I’ve come across scalpers actually working contrary to general lore, and this has sparked my drive to become able to build them myself.

 

Very interesting to hear your takes on this guys.

 

The most impressive one I came across so far is:

 

[spoiler]

7 = O

5 = T

2 = O

1 = R

6 = R

3 = B

4 = O

8 = N

[/spoiler]

 

 

Regarding the negative R:R, I heard interviews with traders on 52Traders.com saying traders miss to include Probability in that formula also. R:R like 1:3 is ideal, but it will also be harder to get high probability/win rate. These factors of course depend on what kind of system one wants to build, just that it’s important to keep in mind the Probability aspect of that formula as well.

 

 

What have you guys found to be the best indicators/signals/trading logic to get close to immediate movement and a few pips to bag? I’m talking about the whole trading universe and all of your trading experience now, not just what’s pre-loaded in SQ neccessarily. Using PriceHigh and PriceLow maybe one of them?

 

 

I generated strats on EURUSD M5 in SQ with many trades (1500 up to 10,000 trades) and I had a few pass all the robustness testing (ZZ’s book + a few extra of mine) + Asirikuy data from 1986. With compounding that kind of number of trades is magical. I haven’t live tested these strats yet, will soon. Do you have experience with robust M5s or have you found them to break in real trading?

 

We have created some M1-strategies on USDJPY,  EURUSD and Cable. Should i upload?

(Removed in retrospect because it was an embarrassing beginner question and didn’t add any value, haha)

Thank you. I would be delighted to learn from your M1-strategies. You could pm me a download link (sprend.com etc.) or post one here or post it in the Extras forum.

I am very interested as well because I tried various settings in SQ but no matter how I tried i am unable to get a strategy with high trade volume and a decent RDD. If someone can share a HFT str file that will be greatly appreciated. Please PM me if you are willing to share.

Seen all your comments, here are some points i want to make:

to build a scalping/HFT algo you can use a small TP + bigger SL to achieve a higher winning %, but you dont have to, i’ve seen few HFT firms and most of them got 30~40 winning %,

that means they are losing small amounts 60 ~ 70 % of the time but when the profitable trades comes they come big,

another point i want to make is that a scalping portfolio always bring better results, you can scalp within a trend and scalp within a non-trending markets,

if you have few robust scalping strategies for both market conditions and working across multiple pairs, you have done it.

you will achieve many benefits just like in a LFT (Low frequency trading) for example: higher trade amounts + lower DD + higher returns + smother equity curve over all.

all depends on what you are aiming at, LFT/HFT and your building/researching skills,

you have to master IS/OOS, WFA+WFO+WFM, MONTECARLO, and other robustness testings to build robust algos, it will come with time, experience, discipline and understanding from failures.

hopefully SQ4 will bring an easier and highly efficient development environment for us, that way most of us will succeed.

M1 is nearly impossible, just use M5 or M15 as a minimum, but also they wont last long unless you will contently re-optimize the strategy,

thats why i’m recommending using M30~H1, and you can even aim for 10~20 PIPS you dont have to aim for 5PIPs.., 10~20 PIPS is still small amount of pips.., its still counted as scalping IMO.. 

I use M1 for all my EA strategies, be it scalper or breakout.  That’s how fast I want my EA to respond to market changes.  If you observe a breakout, the most pips often happened within the 1st couple of minutes.  If I use M5, I “lose” the opportunity to make the most out of a break out. 

As for scalper mode, if I use any time frame above M1 (which I don’t usually do it), then I need to have very sensitive exit plan such as trailing.  

I know not many favour the use of M1 because they haven’t found the “appropriate” indicator or parameters to “eliminate” the noise.  Hence, the use of multi time frame is very essential to eliminate the noise and also reduce those multiple entries that are not “worth” the risk for eg with the use of a higher time frame such as M15, if the M15’s indicator is showing a bull trend and M1 is showing conflicting signals (bear), it’s not worth entering into the trade based on M1’s signal because M1’s trend cycle is not “strong” enough to take such “risk”.  M1’s bear trend won’t last long and hence will only garner insignificant pips gain.  

In another software, they have indicators to search of % change from open prices, such indicator is very useful to search for strategy that enters during big movement where there is significant changes from 1 bar to the next bar etc.  Another indicator is to search for Peak and troughs and then plot these peak and trough (of the trend cycle) and then find which indicators fit into such peaks and trough.  So with such indicator, if we use it on M1 we can search for scalper mode EA and if we use it on H4, we can search for Breakout EA, and if we use it on M15 or M30, we search for ranging EA, etc.

Developing a Scalper Mode EA isn’t easy.  The difficulties lies in 

1.  Identifying the most optimal trend cycle 

– not too small trend cycle that ends up with multiple error of entries

– not too big cycle that defeat the purpose of being a scapler and ended up being a swing trade, etc

2. Eliminating Error of entries cause by “conflicting” signals, in another words Weak Trends

– use appropriate indicators to help you eliminate side way trending/movement 

– volume, CCI, momentum, ADX etc study these indicator and find the most appropriate parameters and indicators that can give you a “clear picture” on the chart.

If you can’t identify or see these trends forming in your MT4 charts, you aren’t able to know exactly what trading rules to apply.  

You need to see these trading rules in your Chart in order to understand how you develop your trading theories.  

A sound trading theory will produce consistent EA that will not only last over a long period of time but is also applicable to all types of instrument.

3. Identifying Optimal entry and exit

– too sensitive, you get more error of entries (it’s a constant balance and trade off between this two components sensitive vs stability (comes laggy respond), the art is to find the optimal balance.

– too laggy, you miss out bulk of the pips and ended up negative because slow entry and slow exit = very little pips or negative returns

In my opinion, aim for high entry accuracy first and then work on a good “dynamic” closing that is capable to reactive to market volatility (hence, personally I don’t favour TP at all).

I  usually don’t recommend developing Scalper EA for beginners because it is very tedious and needs lots of patience and a keen eye to examine the MT4 charts in order to identify the optimal parameters to use and multiple other indicators to eliminate those “error of entries” in weak and conflicting signals.

My favourite EA are the Breakout EA because they are relatively easier to develop, yield bigger profit, suffer less error of entries and easier to identify these breakout patterns (it’s my trade secret, so don’t ask me for the indicators or building blocks etc).  In my opinion, if a trader can’t even identify Big Trend (which is more obvious), how can this person be “astute” enough to pin point with higher accuracy those small and significant trends(and differentiate from those insignificant trends) .  It’s like you need to have a keen eye to see these small significant trends forming “under the magnifying” glass.

Bottom line, developing Scalper EA is not an issue of luck, it is an art and skill to see through all these noises and filter out those significant ones that you are relying on to enter with high accuracy.   Once you can enter with high accuracy, half the battle is won.  Now the other half is finding a “dynamic” closing so that your EA will ride the waves and not exit prematurely due to TP and exit too late due to laggy indicators.  

I usually use both methods for my exit 

1. trailing between 50 to 100 pips and 

2. use an appropriate indicator to exit that is sensitive enough to garner most pips (such as BB is one of the alternatives and other indicators)

3. Break Even and adding winning lots are very important features to have in a scalper mode too.  

4. SL don’t have to be huge, it is usually at 50 pips (4 digits or $500).  

The way to “reduce” the need for huge SL is to develop highly accurate entry rules.  

Once you have high accurate entry rules, the chances of you betting wrong (and therefore need a huge SL) is greatly reduced.

if you have something i have a broker 0 spread 0 commissions very good slippage and execution thanks @keezilla telegram