Day Trading Strategies for Beginners | Section 11, Unit 10 | ATR Scalping Strategy

Quant Trading Strategies & Ideas
1

In [9] -  take_profit_threshold = 0.03

            stop_loss_threshold = 0.01



for time in msft.index:

        .

        .

        stop_loss = entry_price * (1-stop_loss_threshold)

        take_profit = entry_price * (1+take_profit_threshold)



for example the first long entry is 86.16 so stop_loss = 85.29 - take_profit = 88.74

the stock doesn't move that much in the same day

so all the exits are on 15:59… (EOD exit, about one trade per day)



out [10]



Position Entry TimeEntry PriceExit TimeExit PricePnL

0 Long 2018-01-02 10:28:00 86.16 2018-01-02 15:59:00 85.91 -0.284364

1 Long 2018-01-03 09:30:00 86.08 2018-01-03 15:59:00 86.330.215468

2 Long 2018-01-04 09:30:00 86.60 2018-01-04 15:59:00 87.120.485152

3 Long 2018-01-05 09:30:00 87.54 2018-01-05 15:59:00 88.110.534756

4 Long 2018-01-08 09:30:00 88.16 2018-01-08 15:59:00 88.200.004720



Please review, as I understand this strategy should trade many times every day.



Thanks.

 

2

Hi,



Here, the point is we might have multiple entry signals (long or short) throughout the day, but once we are open in a position, we will only exit if our "take-profit" or "stop-loss" signal is satisfied. This doesn't seem to be the case with the selected parameters and hence we are exiting at end of day. But, if you try changing the "take-profit" and "stop-loss levels" you will find that the strategy does exit before the end of day.



I hope this helps.

3

Hi,



The video Section 11, Unit 1 on (1:22) explains scalping is usually 100s of trades a day.

Yet, the notebook example trades no more than 6 times a day and many days it just enters the open and exits the close.



It got me confused about what scalping is all about. 

Should we strive for a strategy that trades frequently in a day? or is this not a factor?



Thanks.

 

 



 

4

Hi,



You are right when you mention that a scalping strategy is a type of day trading strategy where the trader takes multiple trades, even a hundred. However, you should understand that executing a fixed number of trades is not the aim, but executing trades which are more winners than losers. So even if there is one trade in a day but you are generating positive returns, it should be fine.



Hope this helps

5

OK Thanks,

Appreciate if you can answer another question regarding - 

In the article posted on additional reading  section 11 unit 24 - "The Mathematics of Scalping"
The author writes - quote - 

"The strategy relies on two critical factors:
(i) the alpha signal which tells us from moment to moment whether we should prefer to be long or short
(ii) the execution strategy, or “trade expression”"

In the above given example (the notebook included) is the ATR cross formula, the alpha generating - part I?
Does an “alpha generating signal” need to be quantified separately? The term alpha is not clear to me in this context.

Thanks.
6

Hello,



In the context of trading, "Alpha" is the measure of an investment’s performance compared to a certain benchmark. The excess return of the portfolio over the benchmark index is the portfolio’s alpha. If alpha is positive the investment has outperformed its benchmark. Similarly, if it is negative then the portfolio has underperformed. An alpha of 0 would indicate that the portfolio is following the benchmark perfectly.



Hope it helps.

7

Hi,

The ATR scalping example has a 40% win rate.
So, to my best understanding - theoretically, I could enter the market randomly Long or Short with the same stop loss, take profits thresholds and achieve better results. 
Still struggling to understand what's behind this strategy idea.

I understand what Alpha means but not what "Alpha generating signal" means.
I thought maybe it means the signal should generate better than random win rate.

Thanks.
 
8

You are right, an alpha-generating signal refers to a method or indicator that can predict future price movements and generate returns above the market average (positive alpha).



And a 40% win rate means statistically, you can try random entry and exit and you could achieve a better success rate. But this is theoretically speaking.

You know that a win rate of 40% can still beat the market if the "profit per trade" are higher than the "losses per trade"



In a similar manner, a high win rate with low profit per trade might not beat the market returns.



I hope this helps.

9

Hi,

Thanks, it helps, I thought that scalping is about relatively short trades in short time frames but I see now it may have a wider spectrum of possibilities.

However, one thing I do not understand is that -
If we are looking for expansion of volatility on a ‘1min’ time frame, and the results are that many trades last for hours or a full day. (i.e. the volatility expansion does not occur - for example - 60% of the time)

Does that mean the trigger is not valid and discard it? or we can go ahead and trade it, if the exits are profitable? 
 
10

Hi,



If the strategy results in overall profitability despite a high rate of non-immediate volatility expansion, it can still be valid. If the average profit per winning trade is higher than the average loss per losing trade, you can still have a better performance irrespective of the number of times the signal is generated. 



However, if you are strictly looking for a strategy which has to take trades every minute, irrespective of the PnL, then this might not be the right strategy. It will depend on your perspective after all.



Hope this helps.