KAQ: Actively Managed Stats Model (The fund isn’t run by a computer)

© Moneyweb Holdings Limited, 1997-2005

MONEYWEB: Paul van Rensburg joins us now. He’s the portfolio manager of the Kagiso Active Quants unit trust. Paul, you’ve made the number one spot your own, certainly over the one-year performance comparison, when one looks at the general equity unit trusts – which is really the most competitive of the areas. How long ago was this unit trust of yours launched?

PAUL VAN RENSBURG: Alec, the unit trust was launched on the 26th April last year and, yes, we’ve had quite a good run. It went straight up over the top the first three months. It was the top performer first six months. It’s basically stayed at the number one spot since we’ve had 12-month records. Yes, we’ve had quite a good run – made 60% also over the last year.

MONEYWEB: Sixty-one percent, according to Cadiz – we like to follow their figures. There’s a whole bunch of actuaries who do numbers there, and that’s the end of September. So you were on top at the end of June, you’re still on top at the end of September, but you have looked at things slightly differently. Just take us through your whole strategy.

PAUL VAN RENSBURG: Yes. Well, as our name implies, we are a quantitative investment house, but we are an active quantitative investment house. So what we do is we use a multi-factor stock selection model – I know that’s quite a mouthful. But basically we try and use the power of the computer and the power of science, etc, to try and predict – well, we use the model to try and predict which shares are going to outperform over the next month. But I must also hasten to add, there’s also quite a complement of human skill involved, to be able to understand the model, interpret it correctly, and know when it will give the incorrect signals.

MONEYWEB: Would you ever go directly against what the model is throwing out to you?

PAUL VAN RENSBURG: In certain cases, yes. It’s not a black box. I can see clearly what factors are driving the prediction and, for example, there could be a bogus data point, earnings growth could be misleading because it goes from negative values to positive values. In fact, no growth number actually makes sense then, and so one would clearly override that kind of signal. There are corporate actions, for example, there is stuff like commodity prices moving, that the model might not be aware of. So human understanding is a key complement to the model. But notwithstanding that, the model has some powerful strengths that really help.

MONEYWEB: Paul, what is interesting here is that although you’ve managed to get to the top of the table since your launch – and many investors will stand back for three months, maybe stand back for six months – we are now 18 months later and you are still on top, beating the likes of Tim Allsop’s Rainmaker fund, etc. Yet you only attracted R15m worth of assets. Now Tim must get that in a day.

PAUL VAN RENSBURG: You’re probably right. Alec, it doesn’t mention the years before that we’ve also had good performance, but maybe more on the institutional side. I’ve been managing funds like this for – it must be getting on for six years or so, and the performance has always been pretty good. I’m not saying it won’t have its odd patches or so. The fund is small and we have made no efforts to advertise it on the retail side at all. We use it as a flagship for our institutional products at Kagiso, and we’ve attracted considerable assets there. I think we’ve, in absolute terms, got one of the fastest-growing firms and we are now at R4.3bn or more, having been about less than a billion a year or so ago. So we are very focused towards institutional clients. But there’s no reason why retail investors can’t also obviously enjoy this type of performance.

MONEYWEB: Well, it’s a unit trust, like any other. Just some of the changes that you made in the September quarter, given your good performance – I think there’ll be a lot of interest in it. But we can start with the winners first. You’ve added in the medical stocks, Medi-Clinic and Netcare for the first time.

PAUL VAN RENSBURG: Yes, you’re quite alert. You will see that they are only very small overweights, and we’ve noticed that they’ve started to show a bit of glitter again. They did really well about two years ago, and have sort of faded a bit, with a bit of regulation that was maybe a bit adverse towards it, etc. But we are not strong buyers of those particular shares.

MONEYWEB: All right, well, what are you a strong buyer of?

PAUL VAN RENSBURG: Well, I will tell you what we are the most overweight, and probably for the retail investor and your man in the street, they could probably partake of some of these smaller shares. At the moment, we’ve got overweights in things like Busby and PSG and Sasfin and Digicore. I think we’re the first fund to actually go overweight in Digicore – the old Digicore. Other small firms might be Scharrig Mining.

MONEYWEB: I was going to ask you about Scharrig Mining, because that’s been a top performer.

PAUL VAN RENSBURG: Yes. I mean, we really like it, the fundamental story seems quite good, its price momentum, its  valuation. And it’s not that exposed to the vagaries of the rand.

MONEYWEB: Paul, what about Mustek? I got the annual report today, I’m looking forward to reading through it tonight. I see you have added it to your portfolio for the first time.

PAUL VAN RENSBURG: Yes, we did. I haven’t read the latest results, so I’m not that sure about it. I wouldn’t be a confidant overweight in Mustek at the moment. But I speak from a position of ignorance on that one. I haven’t read the latest results. And in fact, we weren’t that positive on it in the recent past.\

MONEYWEB: So why do you put it in your portfolio then?

PAUL VAN RENSBURG: No, it’s since been trimmed, and I don’t think you’ve seen it since the trimming. But I haven’t read the latest set of results and clearly, if there is some new information there, we might revised our view. But some of the Digicores and EOHs were our preferable picks from that sector.

MONEYWEB: We’ve got Mark Drewell from Barloworld in the studio. We’ll talk to him in just a moment about quite an important global development that South Africans are involved with, too. But I see in this portfolio in the quarter to end of September, you actually lightened your holding in Barloworld?

PAUL VAN RENSBURG: Yes, and it might not have been the best decision. PPC came up with some good numbers, etc. But the earnings have been downgraded a bit recently there on Barloworld. But that might not have been the best decision.

MONEYWEB: All right. And Mittal Steel – that might be a very good decision, certainly if you believe the economists, saying that the steel prices are going to be under pressure?

PAUL VAN RENSBURG: Yes. We’ve lightened quite considerably there, and I think we got it just in time.

MONEYWEB: And two other industrial stocks that you have added to – Reunert from zero to, well, a position of nearly 1% of the fund. And then Hudaco, which is an interesting one.

PAUL VAN RENSBURG: Yes. I think they both look quite promising. And Reunert we didn’t have, and we missed out on. And I think that looks quite promising for the future.

MONEYWEB: Paul, you're a very active manager when one has a look at the changes that you made in the portfolio.

PAUL VAN RENSBURG: Yes, we definitely strive as hard as we can to outperform and get those numbers up there. We have a small fund that’s quite nimble. But on the institutional side, we also run quite pretty large funds as well. But we definitely do what we can to try and get the performance up.

MONEYWEB: Well, they certainly have, Wayne McCurrie – and this is your game, isn’t it? Checking out the unit trust performance?

WAYNE McCURRIE: Yes. I’ve know Paul for a long time, and I certainly know of the methodology behind his model, and it’s proven to be reasonably successful over time.

MONEYWEB: Reasonably? Sixty-one per cent in the past year is better than “reasonably”.

WAYNE McCURRIE: I’ve known Paul many years and I’ve worked with him for a number of years. And it has performed reasonably well over a number of years, it really has. Don’t expect it to show twice the market next year again – that would be very unusual. But I think it’s a logical methodology that they use in quantitative share selection.