In-House Asset Management

Gustaf Hagerud, Deputy CEO and Head of Asset Management, Third National Swedish Pension Fund (AP3).
We have had in-house equity management since the fund started in 2001. But then we primarily had exposure in the Swedish equity market, since we have a large exposure to that market. Internationally, in Europe, which is close to us, we have had both internal and external managers. In 2009, when we really decided to push for a strategy of increasing our index portfolios, we moved more management in-house. In this Alpha-Beta separation strategy all our large exposures in both the domestic and international equity markets were moved into low tracking error portfolios. These Beta-portfolios are either index tracking mandates or index-plus mandates. This is now the case for our holdings in European, North-American and Asian equity markets. European and North American Beta-portfolios are now run both internally and externally. In Asia, however, all exposure is taken within external index mandates. The time difference and the lack of detailed knowledge of Asian markets is the main argument for this decision.
If we have a system to manage the portfolio in-house, we can do it. And then it’s a matter of the performance we can get or expect to get in the portfolio, and our possibilities to actually access the markets. And, of course, we have large challenges in accessing emerging markets. But in Asia since it’s an index exposure, with no active decision-making, we could actually run this internally, but we don’t have to. And, of course, it’s a matter of just looking after the portfolio, understanding what kind of securities to run in the portfolio. So then we would rather use external managers.
The second stage of the Alpha-Beta separation strategy was to move all active portfolio management from traditionally actively managed long-only equity mandates, into internal and external absolute return mandates. The internal active management is now done in long-short portfolios with only risk allocation and no capital allocated. External active management is in the form of hedge funds.
One important reason to implement the Alpha-Beta separation strategy was a more efficient use of fees to external management. We have a certain amount that we can spend on external managers and we’d rather spend this on external managers that can give an absolute return, rather than on those giving us an index return.
Another motivation for the Alpha-Beta separation is that in-house management of Beta-portfolios simplifies quite a lot of what we do in our portfolio, in the sense that we can control exactly what we do with individual securities.
The right tools
It is the proliferation of modern platforms that has given us the possibility to have much more of our asset management in-house. In 2009 we started a public procurement for an order management system that could improve our management of our internal Alpha and Beta portfolios. Our aim was to find an order management system that could be used for multiple asset classes, including fixed income, FX and equities. We also required a system where portfolio trades could be initiated by our internal Alpha-mandates and thereafter be taken care of by our execution desk. When the trades have been executed, they should go directly to our front-office and back-office portfolio systems.
It’s the order management and execution systems that makes this possible. We manage the portfolios electronically. The portfolio managers upload the portfolio every day from our back-office system, and they can execute directly through a broker or through direct market access. We use Bloomberg AIM for this.
It’s extremely easy for us. If it’s our European index portfolio for instance, we have our holdings in the system and we can see the index weights in our benchmark MSCI Europe. From the portfolio view in the system we can automatically create the trades that we want and after execution the trades will go through into our portfolio system. It is an extremely simple solution that we have.
We have only a few people doing the executions, as we don’t have enough resources to have an efficient algo desk. The execution desk managers take part in the management of our large internal index portfolios, and they also execute on behalf of our absolute return managers. The absolute return managers send their trades to the execution desk automatically through Bloomberg AIM.
We have used ETFs and still do, but to a smaller extent. For a large institutional buyer, it’s not efficient to use ETFs, since they are not cost-efficient for us. We do not find ETFs to be an implementation vehicle for any long-term allocation.
The role of external managers
I think we’re moving in the direction of increasingly specialist external managers. But then, if you look at the biggest asset managers in the world, they often have the possibility to track an index better than we have. It’s a matter of resources and research. If we are increasing our internal exposure to the European market, which we have done, then we increase our positions step by step and make sure that we don’t underperform, relative to the index and also relative to external index managers. An internal index mandate in a large liquid market should never be a drag on the return of our total portfolio.
Then you have enhanced or index plus managers and I think there’s still a role for these to be external, and then there are other managers that use sophisticated quant strategies to add some extra return on the portfolio. But, as I said, we try to separate Alpha from Beta.
We don’t have any high tracking error beta portfolios run externally. The large external managers we have are really the ones doing long-short strategies in hedge funds.
We have a similar situation when it comes to fixed income. But, in that case, we’re pushing for more internal management. But you must remember that a fixed income index mandate is very different from an equity index mandate. A fixed income index portfolio is much more complex than an index portfolio on the equity side. This is because on the equity side you can actually hold all the securities in the index, but that’s not the case in the fixed income markets. So you get a larger dispersion from the index and you’re forced to have a slightly larger tracking error. Then there are certain sections of the fixed income markets, such as high yield in the US, which we don’t have the staff to run, and this hasn’t anything to do with whether we have a good execution system or not. It’s a matter of the research on the companies issuing the bonds.
Other pension funds
For us it all depends on our Alpha-Beta separation strategy that makes this possible. Other institutions and pension funds have a very different strategy, with large internally managed active portfolios. And that’s not a road that we have followed. I think, generally, those who have similar portfolios and strategies to us, something like an Alpha-Beta separation strategy, are trying to do what we are doing, because you can implement the strategy without hiring a lot of people.
We are moving slowly and we took a first step in 2009 to implement our strategy; it’s really a steady stretch of processing. And the goal is to have all liquid assets in one system with straight through processing. We’re not there now, but I guess we could be there in five years’ time. I am certain that there will be more sophisticated systems around then that can help across all asset markets.