In May the Johannesburg Stock Exchange (JSE) announced new rules designed to
ensure more stringent oversight of how brokers and trading firms deploy “algo-
trading” techniques on the exchange.
Algorithmic trading involves the use of technology such as computer programs or
artificial intelligence (AI) to allow traders to buy and sell financial assets
automatically, according to predefined rules.
Rather than relying on humans to watch over the market and manually place
orders, which means potentially missing out on strong market opportunities, algo-
trading allows firms to take advantage of beneficial prices automatically when
they become available.
The proliferation of this tec
hnology has been welcomed by some, with one
analysis in the American Finance Association’s Journal of Finance noting that “by
reducing the frictions and costs of trading, technology has the potential to enable
more efficient risk sharing, facilitate hedging, improve liquidity and make prices
more efficient.”
Indeed, algo-trading has become increasingly prominent in recent years as traders
seek to leverage new technology to make higher profits: some industry estimates
suggest that up to 75% of global equity trading is now conducted by algorithms.
While JSE-specific data is not available, academics from the University of the
Witwatersrand have noted that “in the last two decades, this new technology has
gained traction globally and now accounts for the majority of the trading volume
on the JSE.”
New rules for robots
In this context, the JSE is seeking to strengthen its oversight of algorithmic trading
by forcing firms to impose mandatory risk controls for algo-traders: brokers and
trading houses using algorithms must have robust pre-trade controls in place to
mitigate the risk of erroneous orders.
Firms must also ensure any algorithms they use are properly designed, tested,
monitored and subject to consistent international oversight – with accountability
for this resting with senior management.
Furthermore, brokers that allow clients to connect directly to the exchange must
now implement appropriate controls and remain ultimately responsible for their
clients’ algo-trading activity.
These rules have been elevated from “technical directives” into the JSE’s formal
rulebook, which also gives the exchange greater powers to oversee algo-trading
activity and crackdown on any irresponsible behaviour. The changes bring the JSE
in line with international best practices, with the new controls similar to those
which are already in place in markets such as the United States and Europe.
Things can go wrong
M’khuzo Mwachande, an investment banker based in Cape Town, tells African
Business that these new rules have come about as “regulators are increasingly focused on operational and market integrity risks, including erroneous or “fat finger” orders, runaway algorithms generating excessive order flow, spoofing
[placing orders with no intention of executing them], layering [creating a false
impression of supply or demand], quote stuffing [flooding the market with orders
and cancellations], artificial price movements and false appearances of liquidity or
trading activity.”
However, he emphasises that “the core objective [of the JSE reforms] is not to
stop algorithmic trading.
“It is to ensure that if a rogue algorithm, software malfunction or poorly
supervised trading strategy causes disruption, there is a clearly accountable party
responsible for managing that risk.”
Global markets exposure
Japheth Munywoki, CEO at Johannesburg-based investment banking firm
Goodson Capital Partners, tells African Business that algo-trading is currently
more prominent on the JSE than other African exchanges – making it particularly
important the regulator takes steps to mitigate potential risks.
He says the fact that the JSE is home to around 135 cross-listed companies – firms
that are also listed on exchanges in New York, London, or Toronto – means that
Johannesburg is more exposed to spillover effects from algo-trading in those
larger global markets.
“The key issue is that we are not a market that operates in isolation,” he says.
“Since we have cross-listings with other exchanges, then we need to manage the
risks that arise from other markets.
“The JSE is much more exposed than other markets in Africa as it is a bigger, more
developed and more efficient market as opposed to the rest of the continent,”
Munywoki adds. “That is why it is so important the JSE has brought itself into line
with international best practices on managing algo-related risks.”
While algo-trading is not yet as common on Africa’s other major stock exchanges
– hence the JSE being the first African exchange to bring in specific rules around
algo-trading – the rise of this new technology could potentially pose challenges
for regulators and market participants across the continent.
Vulnerable to irresponsible trading
As African stock exchanges are much less liquid than global markets – shares
worth several hundred billion dollars can be traded in a day on the New York
exchanges, compared to a reported typical $1.4bn a day in Johannesburg, for
example – they are more vulnerable to irresponsible trading.
Mwachande explains that “developed markets such as those in the US and Europe
have very deep liquidity pools, large numbers of participants, sophisticated
market-makers and extensive surveillance infrastructure. While algorithmic
trading dominates trading activity in those markets, there are also multiple layers
of controls designed to absorb shocks… Many African exchanges, in contrast,
have thinner trading volumes, fewer active participants and wider bid-offer
spreads,” he adds. “In such environments, a malfunctioning algorithm or poorly
controlled trading strategy can have a disproportionately large effect on prices
and liquidity.”
“The JSE faces the greatest exposure at the moment simply because it is Africa’s
most technologically advanced and liquid market and attracts the largest
institutional and algorithmic trading flows,” Mwachande tells African Business.
“However, smaller exchanges may actually be more vulnerable if algorithmic
participation increases because they generally have fewer safeguards, less
surveillance capacity and significantly lower liquidity.”
Can regulation keep up with reality?
As financial markets across the continent continue to modernise, ensuring that
regulation and oversight is updated at the same rate as technological change will
be vital for the JSE and other African markets.
Munywoki notes that “the key issue is making sure that we do not stifle
innovation in terms of technology for trading. Algo-trading has been an important
innovation for improving liquidity and price discovery, as well as securing tighter
spreads. We need to encourage that while protecting investors and market
integrity.”
Mwachande believes that this is not a binary choice: “in fact, robust controls can
encourage innovation because institutional investors and market participants are
more willing to deploy advanced trading technology when they have confidence
in the integrity and resilience of the market.”
“The broader story is that African capital markets are becoming more
sophisticated. The presence of algorithmic trading is actually a sign of market
maturation. The challenge for regulators is not to prevent technological progress,
but to ensure that market infrastructure, oversight and risk controls evolve at the
same pace as the technology being deployed.”