Algorithms have a very niche appeal in the Indian market. Typically, investors think that algorithmic investing means high-frequency trading – which could not be further from the truth. The truth really is that an algorithm is a set of rules that a computer can follow to make investments. The term of these investments can be anything. It is true that an algorithm will have a tough time picking a portfolio for a 20-year horizon, but algorithms do fairly well in long-run portfolios where rebalancing is done quarterly, half-yearly or annually.
Rebalancing gives algo the ability to adapt
Algorithms are typically picking a portfolio based on quantitative factors (based on either market prices or fundamentals or factors) which provide new inputs every day in some cases every minute or even every second. Repeated trading in or out of an asset will ruin returns. So, the algorithm developer will typically set up a rebalancing frequency which brings your portfolio back on track.
An algorithm is a computer program and so does not have any problems of misaligned incentives. Incentives can be misaligned in the case where you are compensating your advisor based on distribution incentives or brokerage. An algorithm only works on data, and thus, works in your interest.
Algorithms can analyse a wider selection
Typically, an algorithm can work with terabytes of data in fractions of seconds relative to the human fund manager or advisor. Thus, it can provide a mathematically sound portfolio from a much wider array of financial instruments. As algos can run 24 hours a day they can also alert you faster of portfolio events through the data. For example, an algorithm can enable you to find the highest quality stocks with certain parameters from a set of thousands of companies versus a human would take months to run the same analysis.
Algorithms provide an unbiased view of the situation at all times. The human brain is impacted by emotional events which are bound to happen in anyone’s life. If your advisor’s day is going well, he is likely to be bullish and vice versa. In the case of an algo, it will only act based on the data and not on external emotional events.
Avoiding repetition of mistakes
Algorithms can be programmed to avoid mistakes that have been made in the past. They can also be programmed to follow advice that you believe should always be followed. Some market events may happen once in a decade which may be missed by a human advisor due to the rarity of the occurrence, but once the algorithm has the feature it is unlikely to miss the event.
You should study algorithm and the firm offering the algorithm in detail before committing your money. In addition, always seek a back-tested set of results, portfolios, graphs, and analytics to make sure that the results are genuine. All retail algorithms for long term investing are non-discretionary to protect the investor. You should always check the output of the algorithm before trading with the output in the market.
The writer is CEO & co-founder, Wixifi