The dynamics of the relationship between the traders and portfolio managers has always been a source of fascination to me. This is partly because trading has only relatively recently emerged out of the portfolio management skillset, but it also derives from a residual sense of ‘them and us’ between the two functions.
Over the last 10 years, the application of process to trading has been a major theme. Traders that were once given little discretion for enterprise or innovation by their portfolio managers, gradually turned more and more to machines, ironically sometimes granting their algorithms more autonomy than they had once enjoyed. For global long-only asset managers running funds for multiple clients scale is inherent to the model. Because the majority of their transactions by volume take place in the equities markets, asset managers were quick to take advantage of the industrialisation of share dealing undertaken by their sell-side counterparts in league with exchange operators. The development of centralised trading desks was largely an equities or equity-led affair with trading in other asset classes added later or not, depending only in part on volume.
High-volume markets with multiple types of participant, such as equities and FX, inevitably have lower margins, the former the more so by being exchange traded and therefore more transparent, but the industrialisation and professionalisation of buy-side equity dealing allowed asset managers to increase their control over execution costs considerably, whether trading directly or through monitoring the performance of brokers acting on their behalf. This greater control over the implementation of the investment process expanded capacity and turnover, thus offering portfolio managers greater scope to adjust their portfolios in pursuit of better performance. It was also a not insignificant factor in allowing successful firms to demonstrate their efficiencies and enhance their yields, thus growing their assets under management.
Industrialisation is synonymous with specialisation, automation and commoditisation. Have buy-side traders automated themselves out of a job? Have they specialised themselves into isolation? Is trading in danger of becoming a commodity that can be outsourced like payroll or HR? A rapid end to in-house trading desks seems unlikely in a post-crisis, mid-reform environment in which all financial market transactions and related processes must stand up to ever higher levels of scrutiny or forensics, while asset classes currently less liquid and less standardised than equities are being forced by regulators to take more equity-like forms, making them ripe for use of tools such as algorithms and transaction cost analysis. In some respects, the plate of the buy-side trader has never been fuller, nor his role in the investment process more important.
The matching of human and machine-based skills seems to be integral to effective trading. In some ways this combination is better established than further upstream in the investment process. Since they sloughed off trading responsibilities, portfolio managers have not yet been challenged to identify which bits of their skillset are better outsourced to machines in the same way as traders have. But surely that day is coming closer. In fact, are smart beta products the portfolio manager’s equivalent to algos? They don’t do the job with quite the finesse and the judgement of the human, but they are considerably more reliable and low maintenance in the long term. Whether the mutual threat of automation brings traders and portfolio managers together remains to be seen.