Where Is The Algorithmic Trading ?

As the S&P 500 continued to plot a relatively sedate upward trend during April, I noticed again that the financial news providers weren’t talking about algorithmic trading. I can guarantee though, that the impact of algorithms on market moves will become a hot topic again as soon as we experience a strong downward trend. 

The fact of the matter is that the algorithms continue to trade actively, irrespective of market conditions. In the highly liquid US and other developed equity and derivative markets, the majority of trading volume is executed electronically. Market makers stand ready to buy and sell continuously during market hours. 

Market making strategies tend to have the shortest holding periods, as the market maker aims to hedge the risk as soon as possible by closing the position and capturing the spread. This spread is generally very small, often 1 basis point in highly liquid equity markets. In times of increased volatility, the spread will increase but the market maker is subject to increased risk while holding the position. For less liquid positions, the holding period of the strategy could be overnight to a few days. In order to compensate the market maker for the risk associated with a longer holding period, the spread will be higher, often in the 5 to 8 basis point range. Spreads on these trades will also be higher in periods of increased volatility as the market maker is compensated for the increased risk. 

 In practice, market making strategies often build positions that have to be recycled. Ideally, the market maker will have access to an index trading desk that can provide a choice price (i.e. the index desk will not charge a spread), for an accumulated position and can recycle the risk. Index desks make markets in products such as futures and ETF’s so are highly proficient at trading large volumes of the underlying assets, as they actively hedge their index positions. Index desks also constantly manage the so-called lumps, the positions that are either greater or less than their relative weighting in the decomposed index. This allows the index desk to efficiently manage concentrated positions accumulated by the market maker. 

Generally, market makers are more profitable in times of increased volumes and volatility. For this reason, the Bloomberg VOVO indicators are fairly strongly correlated to the returns of market making businesses. But, as market making strategies are generally passive, it’s hard to argue that these algorithms contribute to increased volatility. 

In modern electronic markets, algorithms are used to carry out normal trading functions. There are many different types of algorithms and I have focused on market making here but in practice all types of algorithms continue to trade, irrespective of market conditions. Trading applications can process an enormous amount of data, which improves decision making and price discovery. The algorithms trade continuously but it seems the financial press only focus on them when looking for a scapegoat for large downside moves.