Erik Vynckier, Member of the Board of Directors at Foresters Friendly Society, gave us an in-depth insight late Monday morning on Systematic Risk Factors and How to Identify Returns.
The investment landscape today had changed dramatically, he began. In an era of low rates, classical fixed income was very unattractive, while investment and returns were now truly global; international diversification didn’t work as well as it once did.
The last few years had been dominated by increasing regulatory frameworks, and passive investing had become much more popular, expanding into many other asset classes.
“Risk factors are a debate as to what people are really remunerated for by investing in”, he explained, adding that we had gone from traditional beta – à la Markowitz and Sharpe – to smart beta.
Risk factors included Risk Sharing, Structural Constraints and Information Processing, where compensation is given in return for bearing risks, having fewer investment constraints, and processing information, respectively.
But some pertinent questions to ask about risk included: What is the evidence for risk factors? Is there an economic or behavioural rationale? How about data mining and survivorship bias? Are risk factors long lasting? Then there was the issue of smart beta and the passive-versus-active debate, as well as the question of style exposures in passive benchmarks and style drift in active benchmarks.
Erik then outlined a timeline of academic studies. Where Markowitz had a fairly basic quadratic programming approach, Sharpe and Lintner gave a narrative around it. Ross asked why we were doing simple linear regression, when we could be doing a multiple linear regression. He formulated the theory while Fama and French put it into use. The first to look at econometrics intensively were Jegadeesh-Titman, who identified the momentum factor, and illiquidity was seen to also be a systematic factor by Pastor-Stambaugh.
Many strategies appear in a wide variety of asset classes, said Erik, and you can build a diversified portfolio.
How you construct portfolios
“Quadratic programming model had been very influential although unrealistic, and it had been the theoretical justification for passive investing,” stated Erik.
But some people quickly found that you could systematically outperform the classical market-cap weighted benchmark, through equal weight in each security, risk parity and minimum correlation, to name but a few, he added.
Erik went on to discuss optimising over sets of scenarios, globally across different asset classes and currencies, admitting that an efficient way of implementing it was often through passive approaches. “I’ve been a bit critical of passive, but they are a useful tool.”
Pitfalls of Active and Passive Investing
Erik then outlined the pitfalls of active and passive investing.
“Ex-post active investment is a zero-sum game, and on top of that investment skill seems not to be consistent,” he explained. “We could question, for instance, whether there is outperformance at all or whether there was survivorship bias; investment skill could be a camouflage of risk factors.
“We should also question whether the active manager is sticking to certain factor exposure or whether he admits style drift,” he added.
Other pertinent questions to ask included whether you could find the right people, invest before the skill has faded, and get out in the case of underperformance.
But you can start criticising passive likewise, said Erik. “You can say passive trackers, particularly the cap weighted benchmarks, invest most in the most overvalued companies.
“Passive trackers are really investing in late trends, in trends that are towards the peak, if you miss that early trend is it worth doing it at all or are you just investing at peak?
“Passive benchmarks carry hidden risks, and those risks are very often uncontrolled.
“And if you only invest in ETFs you’re going to miss the outperformance of illiquid assets,” he said.
Big data & investech
Erik concluded his presentation with a look to the future. The was promise in big data, and he was doing a lot of work under the headings of data science in finance; complex networks in finance; financial econometrics with high-frequency data and news; as well as big data applications in finance. A lot of it was looking to expand data set and maybe understand more on smart beta, said Erik.
Finally he touched on the inevitability of InvesTech, and how perhaps the biggest risk for big banks was the 360 degree competition they now had to contend with.