From our vantage point in 2039, we look back on the advances that twenty years of technology have brought to the buy-side.
Looking back at the advances of the last 20 years: how the world has changed! No private cars, and clean energy technology everywhere. Under the Great Warming, rapidly rising sea levels have inconvenienced millions, meat and fish are now an occasional luxury, and the weather is definitely more extreme (despite the repeated efforts of the now marginalised Republican party to claim otherwise).
But there are compensations. The extensive vineyards in Surrey’s Mole Valley are now producing a Grand Cru as good as any from France, without the punitive post-Brexit import duties, and the algorithmically generated pop that was popular in the 20’s has thankfully disappeared.
Let’s take a retrospective look at the buy-side. Moore’s law and network speeds have continued to govern increases in computing speed, but advances have been quantitative rather than qualitative. Quantum computing continues to hold promise, as does quantum encryption, and has revolutionised other areas of AI, but has had little effect on the buy-side.
The sell-side began pulling out of analytics provision around 2015 with the sale of Barclays Point and UBS Credit Delta, when it became clear that the soft dollar market was unprofitable, leaving the coast clear for much more nimble boutique vendors. Widely used systems from the early years of the century were sold off to other vendors to acquire their data businesses, while marginalising or closing down the analytics side. This looks inevitable now, but took many at the time by surprise, and caused a great deal of annoyance. It took a few years for the dust to settle, but provided the conditions for numerous new boutique vendors to thrive.
Reporting AI and machine learning have finally come of age. PMAR now looks at how the markets have moved, where the portfolio is affected by these movements, and what the effect is - and writes the result in English, Mandarin, or whatever media you prefer.
Securities and attribution models remain, curiously, largely unchanged. Apart from the advance of new security types, such as weather and global warming derivatives, the conventional securities are still there - with a few additions; the advantages of longevity derivatives and tontines - roughly equivalent to annuities for groups - are now well understood and widely used, while analytics such as curve decomposition, credit spread, and duration allocation remain unchanged. Central banks even publish population/age graphs together with other banking statistics.
Increasing standardisation has meant that risks can be very carefully tailored. For instance, the new CBOT curve twist future allows exposure to exactly that sort of risk and no other - unlike old-fashioned treasury notes and corporate bonds, which provided a basket of risks in each one instrument, not all of them welcome.
After the initial hype subsided, blockchain rapidly found its way into the derivatives market, but not into futures, where a central repository for margin is still needed. Never before has the City motto ’My word is my bond’ been more appropriate.
Data management has improved beyond all measure, and would be unrecognisable from even 20 years ago. Intelligent agents now trawl through databases, asking questions and categorising data so that old-fashioned SQL queries are as obsolete as COBOL.
Instead, one shows a file with metadata, expresses questions as English, and leaves the intelligent database to provide what’s required. Most of the identification of columns is now automatic, and the systems are intelligent enough to recognise most fields and categorise file structures as required. In practice, this means systems are now genuinely interoperable.
In one of the less predicted movements, benchmarks have disappeared and the extortionate fees charged by banks dried up, since a combination of torrenting and anonymity made it straightforward for all market participants to share a common view of the market. Curiously, the experience of the ’dark web’ and torrenting of movies and CDs carried across very easily to distribution of benchmarks.
With its roots in the open source movement, free benchmarks are now maintained by a few experts who are renumerated via micropayments. New benchmarks become available via Decentralized Autonomous Organization (DAO) approach, another technology that depends on blockchain, so that market consensus is quickly reached.
Looking back, we can see that what held the the buy-side back wasn’t lack of computer power, but insight into what existing technologies could do for the industry.
A note to the reader: This highly subjective view will certainly prove to be wrong, but it may provide an interesting insight into the current state of the art.