A portfolio manager friend recently confessed to me his view that learning to code was the one key skill that would help take his career to the next level.
It struck a chord as I chatted to this week’s interviewee, Hoshang Daroga, quantitative investment manager at discretionary fund manager Copia.
Hoshang is a rare example of a techie-turned-portfolio manager. The former computer programmer believes his skillset and past experience transfers well to his current role, because investing is really all about logic and consistency.
Quant investing is undoubtedly gaining traction. Robo-advice and quant investment strategies have become popular in the US, although we haven’t yet embraced this approach with quite as much enthusiasm in the UK.
Hoshang’s interest in investment management was, in fact, first piqued when he was working as a research analyst at a US investment bank supplying research to fund managers. He drew two conclusions:
- Asset managers were moving towards quant investing because the usual run of active managers couldn’t outperform the benchmark. More and more firms were trying to build a systematic way of investing in markets.
- Those asset managers who were successful invested in tech and upgraded it year on year. BlackRock managed $1 trn in 2008, they manage $6 trn now.
By his own admission, it was a baptism of fire: “I joined TransMarket Group as a proprietary trader in one of the biggest credit events in financial market history,” he says. It meant being adaptable in rapidly changing market conditions. “You have to develop strategies quickly. We had to take advantage of the volatility and become outcome-oriented.”
2015-present: Portfolio manager, Copia Capital Management
2013-2014: Masters in Financial Engineering, UCLA, Los Angeles
2012-2013: Equity research analyst – investment banking, MLV & Co
2010-2012: MBA, Fordham University, New York
2008-2010: Senior trader, TransMarket Group
2006-2008: Programmer, Tata Consultancy Services
Trading is a different discipline from portfolio management. Quant investment strategies tend to work well if behaviours are predictable. In Hoshang’s view, this is why most quant asset allocation buys into passives; they behave predictably and consistently.
“In an ETF we know what we are holding, what its delivery will be and how it will react,” he says. Discretionary managers have shifted gear in recent years; the bespoke service that had long been their bread and butter now co-exists with model portfolios where a standard range of risk-rated portfolios can be bought off the shelf. Copia, the DFM where Hoshang is now based, is taking a different and more tech-driven approach. It is using computer models to drive customisation: marrying modern methods with the traditional hallmarks of the discretionary manager.
Copia’s approach is not quite bespoke. But it is targeted at specific client groups and it requires a good understanding of their objectives.
Advice firms give Copia a specific mandate based on their clients’ needs. Hoshang says that by building automated computer models, Copia can offer these clients customisation without the eye-watering charges that would be incurred by employing a raft of clever analysts. In fact, Copia employs a lean five people. Hoshang is bringing to retail a slice of the institutional world, where he learnt his trade.
Setting up the model is the complicated bit, but once it is up and running it will consistently manage the client objectives. It manages risk by analysing enormous amounts of data. “There are certain pockets of the market that can be harmful,” says Hoshang.
Data also drives performance and the availability of data is forever changing. As new sources emerge, the computer models incorporate the new data to try to improve the accuracy of their predictions. A recent example is sucking in live data newly available from Chinese companies.
Hoshang likes the fact his models remove human bias: “When I was a trader I’d look at the discretionary traders. Sometimes they were right, but sometimes their egos got in the way. So, remove the bias about what is happening in the market.”
As investment horizons lengthen as a result of the pension freedoms, he also sees the value of the consistency that a quant approach drives – as well as the customisation it can offer at a fraction of the cost of traditional methods. He believes this will help advisers stand up to regulatory pressure to demonstrate suitability and good client outcomes over the long term.
If Hoshang and my coding enthusiast friend are right, marrying computer science and asset management will become mainstream. Fund managers behave subjectively in the face of market events but computer models are consistent. This could lead to a big shakeout in the market, Hoshang believes.
To many of the newbies entering our industry writing code may be as natural as learning another language. The next generation will apply its own “digital native” logic to managing investments.
I’m off to familiarise myself with the basics of Java script – no doubt already obsolete. This Luddite, for one, is looking forward to seeing what this generation can do with my investments.
Miranda Seath is research director at Platforum