Type of strategies - understanding the possibilities

Types of strategies - understanding the possibilities

Going deep into the fantastic world of algorithmic trading can be breathtaking, yet frustrating at the same time. There are many different types of trading strategies to develop, especially for newbies. It can lead to confusion, and it can be hard to decide where one should lead his direction.

To make it easier, here is a summary of different algorithmic trading strategies that exist.

stress tests

A] Mean Reversion Strategy

Mean reversion strategies are one of the most commonly used trading strategies in the industry. Their fundamental assumption is that prices (or their ratio) will revert to the average price over time. This omnipresent phenomenon is called the regression, and it is a natural law you can regularly see in your daily life. In the case of trading, if the market is in a significant and fast downtrend in one day, we could estimate that the market won’t be such a considerable downtrend tomorrow. In other words, we expect the next day to move close to the average. A popular alternative to just looking at one security is to look at multiple different securities and their correlation.

This type of mean reversion strategy is known as pairs trading. Trading on stock exchanges worldwide is one of the most profitable and “sexiest” enterprises one can come across. In business terms, relatively low capital is needed for a start.

Every professional trader knows that risk management is the foundation of long-term prosperity. Banks and investment funds worldwide pay generously for their team of experts that has only one task that is probably even more important than returns: to minimize risks, minimize risks, and, once more, to minimize risks. A stable strategy that makes a few percent per year is incomparably better for a professional than aggressive scalping that can make 100% in one month but can also ruin your account within a few minutes.

A myriad of instruments can be traded on exchanges worldwide, from currency pairs to stock, options, to futures contracts. Each instrument has its characteristics, risks, and profit potential. Equities are the most easily understandable for a beginning trader: when I buy stock, I buy a share in the given company. When things go well, the value of the stock goes up. When things turn for the worse, the value of the stock (company) goes down.

There is one well-known mean reversion strategy that takes upsides and downsides moves into account: Stock pair trading – statistical arbitrage.
Banks have looked for reliable and stable exchange trading strategies since the beginning of their existence. Dozens or even hundreds of more or less successful methods have been devised over the years. Some strategies have only worked on a short-term basis; others made money for decades. A strategy that has worked on an extended basis is stock pair trading, also known as statistical arbitrage.

Statistical arbitrage was described in theory in the 1950s by the Australian investor and hedge fund manager Alfred Winslow Jones. But it did not get used in practice until the introduction of powerful computers that made it possible to process enormous volumes of data in real-time and search for profitable equity pairs. The strategy boom came in the 1980s thanks to Gerry Bamberger, Nunzio Tartaglio, and their expert team from Morgan Stanley. In 1987, Morgan Stanley announced that it made 50 million dollars on Tartaglio’s automated systems. That was an incredible profit for that era. Although the group was dissolved in 1989, stock pairs have remained one of the bank’s strategy pillars. Over time, as details of the strategy reached the general public, stock pairs became one of the most popular neutral strategies. Today, they constitute the foundation of trading portfolios of professional traders, banks, and funds worldwide.

The principle of stock pair trading

The concept of pair trading is surprisingly simple: we find two shares whose price ratio has been stable on a sustained basis and speculate whether it will continue to remain stable. In other words: if the price ratio (stable in the long-term) sways outside of the usual range, we speculate on it returning to normal after some time.

Stock pair trading is known as the stock spread, or as the statistical arbitrage – this is because this kind of trading is based on statistical analysis. Stock pair trading is one of the investing strategies used by professional traders, banks, and investment funds. Stock pair trading is a strategy based on trading two different stocks at the same time. The goal is to profit from the price difference between those two stocks. Two stocks are being traded during the stock pair trading. Both stocks have very similar behavior (they are highly correlated). Here is an example: We have two companies on US stock exchange (Figure 46): SMLP Summit Midstream Partners LP (NYSE); ENLC EnLink Midstream LLC (NYSE). Both stocks come from the same sector (Natural GasDistribution). One stock is bought (Long position), and the second one is sold (Short position). Stock pair trading is based on the price difference between two correlated stocks. We can speculate on the price convergence of correlated stocks or their divergence. If we can find a stock pair whose price ratio is stable in the long-term, we can speculate (with a high probability of success) on returning those prices to normal.

Figure 46: Price chart of highly correlated stocks

Figure 46: Price chart of highly correlated stocks

When trading stock pairs, we do not speculate on the fluctuation in the price of one stock, but on the relative fluctuation of the prices of two stocks concerning one another – the return of a short-term sway in their price ratio back to the long-term normal.

In practice, a pair transaction works as follows: at one time, we buy the undervalued stock title and sell the overvalued. If the assumption of the reunification of their prices is confirmed, we collect the profit. It is irrelevant whether the profit is generated by the growth of the undervalued stock title, a drop in the overvalued one, or any other combination of price fluctuations. What matters is that we collect our profit when the difference in prices is reduced.
For greater clarity, we show a figure that demonstrates the principle of trading on the difference in stock prices in simplified terms. The difference in stock prices, in this case, is expressed in simplified terms as a simple difference in the prices of stock the titles used (there are several ways of expressing that difference).

Again, we are using SMLP and ENLC stock for our example (Figure 47):

Figure 47: The principle of pair trading

Figure 47: The principle of pair trading

  • Point 1: The difference between stock prices is very small
  • Point 2: The difference between prices is unusually high – you can expect a return to the long-term average
  • Point 3: The prices became closer again
Trading automation

Pair trading is simple in principle, but it requires very high gross calculation capacity. At one time, one must monitor dozens or hundreds of stock pairs, assess their trading model, and execute the relevant orders precisely and flawlessly in the event of a signal for entry/exit.

The edge of stock pairs

Tree pillars can make the edge:

1) limited risk of loss,
2) speculation on the return of the ratio of prices to its long-time average, without the influence of external factors,
3) broad portfolio diversification thanks to the many trading opportunities in the stock pairs universe.

If we trade in two stock titles in a pair opposite one another, i.e., one Long and the other Short, the external factors will balance each other. It does not matter whether the entire market moves upwards, downwards, or to the side. The ratio of the prices of two randomly selected (correlating) stocks can be stable in time, oscillating around the long-term average. Using statistical methods, we can study that behavior, describe it, and quantify it. Suppose the ratio of prices sways outside of the usual range (usually 2nd standard deviation). In that case, we can speculate with a high level of success (around 66%) on a return to normal, thereby collecting a profit.

No trading strategy generates profits solely. Loss is an integral part of stock trading, and stock pairs are no exception. A loss occurs when the difference between prices does not revert to normal within the set time. Rules for leaving a position must be defined to deal with this adverse development. Simultaneously, a very effective method is a time stop-loss – closing a position no later than at the end of a day defined in advance.

Based on an analysis of an extensive set of data, it is possible to state that stock pairs manifest a stable winning ratio of around 65%. Risk Reward Ratio can oscillate between 0.65 – 1. Of course, the higher RRR, the better for the performance of the strategy.

By nature, a stock pair is hedged against market fluctuations. It is a market-neutral strategy.

If we combine several stock pairs in a portfolio, the total number of transactions will increase, and the equity could become significantly smoother because of trading many uncorrelated pairs. Of course, this approach needs complete automation, as it is very time demanding.


Next up, let’s look at a different type of algorithmic trading strategy, namely momentum strategies.
Momentum strategies are pretty much the exact opposite of mean reversion strategies. Instead of betting on prices returning to some state, momentum strategies try to profit from continuing a particular move. The underlying assumption for such strategies is that price moves can hold their momentum for an extended period. Let’s go over a few examples to clarify this.
A classic momentum strategy is the so-called “gap and go” strategy. It is a strategy that looks for overnight gaps and is based on a continuation of this move. For instance, if stock APPLE gaps up by 3% overnight, this strategy might seek to establish a shorter-term long position on XYZ. This is exactly the opposite of what a mean reversion strategy would do (there you would bet on reversal move).
Another popular momentum strategy can be based on earnings, or the news event fueled price move. Since significant news events can have a lasting impact on its stock price, you could try to be early enough and bet on continuing this price move. Furthermore, you can also combine this news event with an overnight gap to create a slightly more advanced strategy.


Trend following strategies of the longer-term adaptation of momentum strategies. They try to identify price trends and take advantage of them by following the found trend. Usually, these are longer-term trends, but they can also be mid to short term. The exact definition of a trend depends on the chosen implementation of a trend following algorithm.
Trend-following strategies follow these principles:
It opens many small positions in many different markets and systematically cuts losses as soon as the trend moves against open positions. It means that the system generates more losses than profits, but profits prevail over losses in the long run. Typically, the system’s success rate is much lower than fifty percent.

This strategy stands on two pillars:

  • position sizing, i.e., proper capital allocation among individual markets,
  • trend signals (for example, the crossing of moving averages or a typical breakout signal where we speculate on breaking a certain price level, such as a long-term high or low). 

In most cases, a trend-following strategy doesn’t rely on any fundamental model. It doesn’t focus on a specific asset class, and they are typically based on the long-short model (we open both long and short positions). This strategy is also “agnostic” as to the reasons for market behavior – the system only follows market trends. Therefore, it has nothing to do with fundamental analysis.

Diversify, diversify, diversify

The creation of trend following strategies was possible thanks to the development of information technologies and automated trading systems. However, it’s not too young, and we can evaluate the real trading results from a very long perspective. The results of many hedge funds using trend-following strategies over the last thirty years correspond to theoretical assumptions. Trend following is an asset that shows a positive slope of the yield curve and the potential to bring profits even in market slump times. Drawdowns of trend-following systems are generally shorter and up to one-third lower than drawdowns of buy-and-hold strategies. Drawdowns often appear when the market offers no opportunities and has no clear trend.

There’s one big plus: the trend-following strategy does not behave like any investment asset, i.e., there is no correlation. It makes it a useful tool for portfolio diversification.

Most investment strategies are based on the convergent risk principle, which can generate devastating losses during unexpected market events. Trend-following and similar strategies use the divergent approach to risk and are therefore immune to market drops during crises. This system calculates with volatile growth, and it can generate significant profits under extreme market conditions. Most CTA hedge funds are long term advocates of these diversified trend-following strategies with dynamic portfolio sizing models.

Amongst the most most popular indexes that reflect the performance of the biggest trend-following CTA funds belongs:

  • Altegris 40 Index
  • Barclay BTOP50 Index
  • Barclay CTA Index
  • Barclay Systematic Traders Index
  • CISDM CTA Equal Weighted Index
  • Credit Suisse Managed Futures Hedge
  • Fund Index
  • iSTOXX Efficient Capital Managed
  • Futures 20 Index
  • SG CTA Index
  • SG CTA Trend Index
  • Stark 300 Trader Index
  • Stark Systematic Trader Index

When you look at their performance, you see that the trend-following strategies have not beaten the S&P 500 index over the last ten years. These CTA funds are going through a long term crisis, and it is questionable whether they will be competitive again. In my opinion, the main reason is that these strategies are applied to daily data so that the signal lag could be prolonged in a nowadays dynamic world. Finally, this type of diversified trend following strategy requires significant capital in millions of dollars, and for most retails traders, this is out of the question.

D] Market Making Strategies

Another common type of algorithmic trading strategies is market-making or execution based strategies. Retail traders can’t deploy this type of strategy. But this doesn’t make it entirely irrelevant, which is why I briefly want to cover it. When big institutions such as banks or funds wish to open or close a position, they can’t just buy or sell it on the open market like you, and I would.
Their position sizes are so huge that if they were to sell or buy in one go, they would push the entire market in one direction. Besides exposing their trade to everyone else, this would also lead to unfavorable entry and exit prices for them. It is an especially big problem in a liquid market. So, what institutions do instead is open or close their positions in many different smaller orders. Sadly, dividing your order into many small orders doesn’t completely solve this problem. Since they still might create a lot of pressure in one direction. That’s why they use execution based algorithms that try to find the best possible places to send out more orders without affecting the price too much. Identifying that a big institution is currently using such an execution based algorithm allows you to trade around this algorithm. That’s why market makers often use so-called sniffing algorithms that try to find and profit from big order executions through these algorithms. High-frequency trading strategies usually fall into this category of algorithmic trading strategies.

Advice No.29: Most strategies can be customized so that they are best suited for you and your preferences. Note that this chapter just covered general trading strategies. Regardless of what approach you want to use, it is crucial to implement solid risk management practices. Otherwise, you’re just sitting on a ticking time bomb.

If you don’t want to read all I want to share with you article by article, grab our Ultimate Guide To Successful Algorithmic Trading here and read it anytime you want! 12 chapters, 112 pages: all in one place and completely FREE of charge! 

Leave a comment