Elroy Dimson

I would argue that an active manager that is not particularly persuaded by your sort of, didn't like the passive management, is not persuaded by your interest in factors should still be looking at these attributes.

Yes, although what I'm talking about is traditional active managers primarily who accidentally end up with factor tilts.

And those factor tilts are essentially bets being made through the portfolio where they didn't actually intend

to make those bets.

They were doing something which they thought was more innocuous and more geared towards the objective of the plan.

I've been on a lot of investment committees for pension funds and endowments and I've seen that multiple times over.

But journalists should be aware that if this is a good idea, it would have been a good idea two years, four, six years ago, which would mean that every time any of the magnificent seven do well, you would reduce your exposure.

At the end of the decades, you would feel much poorer.

The size of the equity risk premium is very important.

If you were trying to build a modern building, you will have steel pillars that support it.

But if you were trying to do that in such a way that the ratio of the diameter of those pillars that support a skyscraper, the ratio of the circumference to the diameter, if you wanted to be anything other than 3.14, 159, 635, etc., you can't do it.

So we have a number in investment which is just as important as the equity risk premium.

The trouble is that while we know what the value of Pi is, we have no real idea as to what the equity premium is.

It's even worse than that because there are lots of different estimates that come out.

They seem to vary a lot over time.

We think of these as being long-term attributes, but there are sell-side advisors who are constantly changing their minds and who've got limited opportunity to do their business if the equity premium never changes.

So we see lots and lots of calculations in the academic world.

This has been a source of discussion for 20 years.

So it started out with Goyal and Welsh, Welsh being based in the US, Goyal nowadays being in Switzerland, who looked at what happened if you didn't peek into the future, but just chose, as you went through time, to make an investment based on information that at a particular date you've got based on the past.

The equity premium that you get if you use long-term data is still clear, depending on whether you look at the 21st century, the 20th century, or the 19th century.