Algo Trading Strategist V/S The Manual Trader

There are two types of Algo trading:

1. STATIC ALGOS: The first type allows you to feed in a technical trading strategy, backtest it, deploy it and allow it to trade on your behalf depending on the trading capital you have deposited in your account. Entry and exit points are fixed and the system will merely keep executing trades and reporting profits or losses. Some online brokerages provide this facility, notably Zerodha’s Streak, and let me show you how it works:

I have created a strategy to backtest (results follow). After backtesting (for up to 3 months/CNC (not MIS), and checking profits and losses, I can  choose to deploy it.

The software will then keep trading on my behalf whenever the conditions are fulfilled. All I must to do is sit back and watch and choose to continue with it, or stop it with just one click.

This is the first type of algo trading.

2.  MACHINE ALGOS: The other type of algo trading is more sophisticated because it is based on machine-learning. In the machine learning algos, which are programmed using R Program or Python, the AI self-learns from its mistakes and keeps getting better over time till it becomes fairly accurate.

In this article, the debate is Static Algos v/s The Manual Chartist.

The Manual Chartist

A technical analyst who operates manually typically uses a tool to view charts. He sets alerts based on different conditions, and once he is alerted, he checks the chart and decides whether to place an order.

As compared to the static algo trading, a manual chartist seems slow and clumsy.

But, does that imply that static algo trading is a better form of trading?

Let’s find out:

Static Algo Trading V/S The Manual Chart Analyzer

In the algo strategy stated above, as you can view in the screenshot below, I have allowed the system to buy 100 shares of RIL and SBI when the price fulfills the trading strategy (you can view the strategy in the screenshot). Let’s see how it did:

RIL turned in a profit of Rs 7,626 and SBI lost Rs 390. Net about Rs 7,200. THIS WAS FOR A PERIOD OF 3 MONTHS!

AND BEFORE LEVYING BROKERAGE!!

Let’s see how much we made (in the back test, of course) after brokerage:

 After brokerage, we made abut Rs 5,300 after paying Rs 1,900 brokerage IN 3 MONTHS.

Now let’s see what the manual chartist would have done. I’m just taking his trades for today (5-7-18)

Instead of buying and selling 100 shares, he would have straight away bought 1000 shares of RIL and SBI each.

Based on the same strategy, Reliance gave a buy signal at 9.40 AM when it was 998 and a sell signal at 1002 at 10.15 AM. The trader would have gotten out after 1 just one transaction with a post-brokerage profit of Rs 3,700.

In contrast, SBI did not provide any meaningful signal to the manual chartist. There was a flash at 9.52 AM which may have resulted in a profit of a buck or two or a loss of a buck or two.

However, traders who follow our 15 Minute setup would not have touched it because of falling DI+ and rising DI-.

So, if we applied this strategy in static algo trading, we would have made Rs 5,300 IN THREE MONTHS and had we stuck to manual charting and trading, we would have gained Rs 3,700 (possibly lesser if losses were incurred on the SBI trade) IN ONE DAY!

Obviously, manual trading seems way much better than static algo trading BUT you can tweak and sharpen strategies to improve profits. The conditions are:

1. The strategy has to be top notch and it should avoid whipsaws. For this, you have to learn TA. If you work with a poor strategy, you will lose money.
2. The trades must be continuously monitored and the algo deployment should be stopped after reasonable profits are made (or SLs are hit).
3. The algo trader must have a set of 4-7 strategies that he must deploy during the day at different hours, based on his reading of the market.
4.  The strategy, preferably, should be against the herd (hop to the next section).

How To Profit From The Herd Using Static Algo Trading

You know that there are lakhs of traders who are not savvy on algos and trade on the basis of traditional formulas such as Open=High, Open = Low, etc.

Though it sounds cruel, you can profit from these traditional traders.

What you should do is deploy a traditional formula, for example, Open = High, right at the start. You should do this for a stock that is positive – for example RIL on 5-7-18 because good news was expected in its AGM.

When you do this, and set P/L and SL targets (for example, 2% and o.5% respectively), you should deploy in the first 15-20 minutes and then shut off the system if you have made profits or suffered losses.

Here is an example of OPEN=HIGH strategy (backtest) that I applied on RIL today on 5 Minute candles (as the intention was to make quick and small profits):

In the backtest, we have made pre-brokerage Rs 7450 profits,in double quick time. The profits were made because traditional traders took their time getting in. By the time they got in, we got out.

The minute profits are made, the strategy must be switched off.

This is one example of how you can game the system and this is an effective way of using static algo trading.

I am yet exploring algo platforms in detail and plan to include good strategies in the courses that will start soon.

To conclude, manual charting and trading seems superior to static algo trading. However, you can make money with static algo trading by either working with a super strategy, following all the riders mentioned above, or by gaming the herd (example above). Otherwise it’s no good.

On the other hand, machine learning-based algorithm trading is much superior to static algo trading and manual trading, but that’s another story.