What are Algo Funds?
‘Algo’ is commonly used in short for ‘algorithmic’. An algo fund is one of the many types of investment funds available in Singapore and in countries with more mature trading markets or infrastructures, such as the United States of America and United Kingdom.
Algo funds are active funds that have a software to carry out automated trading. The software follows a defined set of instructions (an algorithm) for placing a trade in order to analyze data and execute trades at a speed and frequency that is impossible for a human trader to perform.
Other terms associated with algorithmic trading are automated trading and black-box trading. There are many subsets of strategies carried out in algo trading. For more details about that, you can read our article ‘What is systematic trading?’
What percentage of trading is algorithmic?
According to data furnished by the Singapore Exchange and published by The Straits Times in 2017, it accounted for 34 percent of the daily stock turnover in Singapore. About two-thirds of trades performed by algos were with institutional investors. The rest were with retail investors.
In 2019, figures reported for machine-led daily moves in US stocks was 80%. Algo trading is now being used in most of the large institutional firms in the American market.
Who uses algo trading?
Institutional investors and big brokerage houses use algo trading mainly to cut down on costs associated with trading. This method of trading can spread out the execution of a larger order, or to perform trades too fast for human traders to react to.
Research suggests it is especially beneficial for large order sizes that may comprise as much as 10% of an overall trading volume.
Algo trading is widely used in hedge funds, mutual funds and pension funds. Thus, a hedge fund traded via algorithm may also be an algo fund, in which case it may sometimes be referred to as an algo hedge fund.
There are many types of algo trading machines. Some are built for institutions and large hedge funds. These usually have sophisticated equations programmed into them, coupled with low latency instructure that are usually beyond the budget of individual traders.
Some are built for retail traders or ordinary traders. These softwares can be purchased for only a few hundred or thousands of dollars. They come ready-made without any need for programming except to input a desired portfolio.
Do banks use algorithmic trading?
Yes, especially investment banks or banks with a dedicated business division for investments. Many of such banks employ algorithms designed to execute trades without significantly impacting market prices, rather than for outright gains.
Is it easy to build your own algo-based system?
Yes, it is relatively easy to start and build one. However, it is hard to have a consistently profitable algo-based system that is scalable as it requires significant expertise, resources and understanding of counterparty relationships.
Here we outlined a summarised process of how one can go about building an effective and profitable algo-based system using an existing ready algo software in the market.
- Find a profitable trading strategy (which in itself is not an easy feat).
- You would have to spend significant time filtering out all the ready-made bad systems to identify those which have the best potential to succeed.
- Define a trading plan, i.e. a set of rules or parameters (including computing variations) for the system to run on. Parameters include any of the indicators, time periods, trading logic combinations, range settings, candle computations, entry and exit rules, price values, order types or custom indicators.
- Back-test the systems according to your trading plan. Other tests include Monte-Carlo testing, Robustness test, Optimization, Walk-Forward testing, demo testing, and live testing.
How long does the entire process entail? The development process (steps 1 to 3) alone takes at least 3-6 months of full time focus followed by one year of live testing.
Including modifications to the system and several rounds of back-testing, the total duration can exceed 2 years before a profitable and robust algorithm can be used.
To top it all off, the market cycle changes every few years, resulting in ‘strategy decay’, where a once profitable and effective algorithm may be less profitable due to market changes.
No trading strategy lasts forever as markets are constantly evolving. Hence, it is important to always review and finetune the strategy and portfolio to suit the current market dynamics.
Is algo trading safe?
As with any kind of investment, there are risks involved. The computer-driven algorithm used is just a tool or mediator for carrying out trades. What matters is how it is used, and that is dependent on the professionals who decide on the parameters and instructional input of the software.
In fact, it may be safer than purely human-run trades. A software is not hampered by human emotions or biasness. It operates entirely on logic and is able to respond immediately to changes in market trends.
Yet fear over machines is a deeply embedded psyche. A paper conducted by the University of Pennsylvania (USA) in 2014 produced evidence of algorithm aversion.
It showed how people are instinctively distrustful towards machines. They preferred to trust human forecasters over algorithmic ones although the latter makes fewer and less severe forecasting errors.
A paper published by two Australian researchers in 2017 revealed the effects of algo trading (AT) on mutual fund performance. They found that ‘funds holding stocks with higher AT intensity have lower holdings return, and higher interim trading profits as measured by return gap.’
The positive effect of AT on return gap ‘survives controls of known factors of fund returns as well as liquidity variables such as effective spread and execution shortfall.’ The paper also stated that non-index funds benefit the most from AT.
Is volatility good for algo funds?
Yes. Although the COVID-19 pandemic caused the Singapore stock market some turmoil, many traders took advantage of the volatility, which led to a surge in algo trading. This was reported in The Edge in July 2020. Evidently, this says that a volatile market can be advantageous for opportunistic traders.
How do algorithms affect the stock market?
There is a lot of critique of algorithms causing volatility in the stock market and affecting liquidity. Recent studies suggest otherwise.
The TRADE’s 2020 Algorithmic Trading Survey was conducted in 2019. Results from the survey showed that more and more hedge funds are using algos mainly to reduce market impact.
Other reasons are the ease of use and ability for algorithms to execute performance consistently. In fact, hedge funds are happy to use algos to trade a majority of their portfolio.
The caveat: the survey was done in 2019 before the pandemic caused the market to fall. Algo trading may possibly cause an escalation or worsening of the situation when a sell-off is triggered by black swan events or the enforcement of new financial rules.
However, it cannot be entirely faulted for a market crash. Any number of factors can contribute to such an occurrence.
A study back in 2014 seems to address the question on algo trades affecting liquidity. According to the study, algorithmic trading makes stock markets more liquid and efficient. Apart from that positive effect, it can also hide the identity of large buyers and sellers.
An important point to note is that many complainants are human traders who cannot compete with the 24-7 capability, speed and accuracy of machines.
A letter to the press written a year later expressed concern that algo trading and large institutional investors with deep pockets will influence prices greatly. Thus, causing a great disadvantage to the investing public
Algo trading and algo funds are here to stay. It is likely that more types of investment funds will migrate from human controlled trades to computer-driven ones.
Since algo trading is able to profit from tiny movements in the market, perhaps one of the best funds to invest in are those which choose to capitalise on FX spot trading, such as the Salzworth Global Currency Fund.
Associate Director, FX Algorithmic Trading
Raphael has a diverse range of skill-sets from managing infrastructure systems to auditing investment algorithms. Since 2013, Raphael has been researching and developing scalable FX Algorithmic Trading strategies that matched global FX liquidity pools.
Leveraging on his Technology and Finance background, Raphael seeks to present a robust set of algorithm trading offerings for Salzworth Global Currency Fund and to provide relevant connections to benefit clients and partners of Salzworth.