Lochiel Crafter, Asia Pacific head of State Street Global Advisors, interviewed the chief investment officer of Sunsuper, David Hartley, about super funds bringing investment management in house. The discussion focused on key drivers as well as what funds need to consider before making the move – and why Sunsuper does not yet undertake significant inhouse investment management.
Lochiel Crafter: What do you think the drivers are for internalising investment management, particularly amongst large superannuation funds?
David Hartley: Many super funds are internalising their investment management function to save cost. However, I think if you’re doing it just for cost there’s a logical disconnect. You are effectively hiring a group of people and paying less than it would cost to get those same people externally.
One of two things is going to happen. They are either going to be unsuccessful, in which case you’ve got all the problems of rectifying that. Or, if they’re successful, you’re paying them less than what they can get in open market. So you have to expect to lose those people. So, as a fund, if you’re doing it for cost, you’re going to have to be prepared for continual turnover.
The other reason funds internalise is because the asset-owning model for some of these assets is better suited to internal management. For example, some large defined-benefit funds in North America have insourced infrastructure investments. One reason for this was the belief that the North American model for owning infrastructure assets was wrong. It had grown out of a private equity model. So you had limited life investment vehicles that bought long, stable cash flows, geared them up and then t took money out after a 10-year period.
However, some defined benefit funds argue, “We want these assets for 50 years, we don’t want them for 10 years. We want them because of the long-term cash flows, which are inflation linked.” If you gear them up, then you have to strip out those inflation-linked cash flows, you have to do swaps and all those sorts of things, so you can get a financing program replaced. The current ownership structure was inconsistent with the investment objective, because it was taking away the very characteristics that investors wanted from those assets. Internalising investment decision making and direct ownership resolves some of these issues.
At the same time, there is a recognition that the decision is not all about cost. Funds need to employ people who were strong managers of those assets and know what they are doing with them.
LC: Why is it that many funds are so focused on the cost saving aspect?
DH: The total fees we’re paying out to external managers are about $100 million a year. If we doubled in size, we would get some benefit of scale, but pretty much it would double the fees under the current model.
$200 million can buy you a fair bit of internal expertise, and that’s when people start saying, “Can I save on these other aspects. Can I beef up another part of the portfolio, so I can get a better value out of the money I’m spending?”
Managers don’t recognise that, as funds get bigger, they can manage money themselves. It doesn’t cost a manager twice as much to manage twice as much money, so why should the fee be close to double?
Managers generally, I think, prefer to have 10, $500 million clients than one $5 billion client, and they reflect that in their fee scales. So as funds grow, they don’t get the full benefit of the scale that they’re generating.
LC: Is there a minimum size at which it starts making sense for a fund to internalise?
DH: It depends on what asset class and investment style you’re talking about. For example, for Australian shares it would actually be more expensive for us to in-source at the moment. We’d have to pay the costs of the staff and the systems, and the governance structure. Plus, with Australian shares we’ve got pretty low marginal fees.
We think that we’d probably have to be about double our size before it starts to become a meaningful cost-benefit, just in terms of the fees. Then there’s the question as to whether we’re giving up return as well, because it’s not just about fees.
LC: What do funds need to think about when considering insourcing?
DH: A fund goes through a series of stages as it grows. I refer to this as ‘punctuated evolution.’
When it first starts up, the fund has just got some sort of governance board. The first thing they do, typically, is hire a consultant to help them.
The next stage is to bring in an investment expert to work with the consultant and focus them on particular areas. That internal person soon says, “I can do more if I can get some more people,” so you start to build up an internal manager-of-managers team. I’ve seen these all around the world in different areas, companies and funds.
Next, they start to say, “We can manage some of this money ourselves internally.” And then, “Maybe we should manage money for other people as well.”
The next phase is when the fund views their internal investment team as just another manager, and they’ve hired people into the trustee office to look after them. The trustee board says, “They’re so far removed from us now we’ve got to have people in-house to manage them.”
When you think about it you’re just going back to a manager-of-managers structure.
LC: How should a fund manage its internal team, in terms of governance, structure and the review process?
DH: It should apply the same governance standards internally that it applies to external managers, particularly for defined-contribution money.
Inevitably, any active manager is going to have a period of under-performance at some stage. You need a consistent process for how you manage that. If you’ve got a separate governance structure you can decide to reduce that function.
But if you say, “We’re committed to internal management of this particular function, and we’re going to internally manage it no matter what,” then the trustee is potentially compromised.
In terms of the structure, the overseas funds I spoke to that have internal functions have set them up as separate companies. This is partly for salaries, so that they’ve got a chance of holding on to those people – because they can offer them some sort of shadow equity, or something similar. And partly it’s for governance, so they can have a separate decision as to whether they’re keeping that team for that particular asset class or that particular function.
It makes sense. If you’re going to take on managing money internally, and at some stage you’ll think about managing money externally, you’ve really got to have that as a separate company.
LC: What elements of the investment function do you think are best suited for that insourced function, in terms of asset allocation, manager selection, portfolio structuring, all the way through to ultimate trading?
DH: I think strategic asset allocation is well suited to management in-house. That said, a number of managers are becoming more like consultants and offering multi-sector type mandates. These used to be the norm in Australia up until 1987 – when clients discovered their managers were doing things they didn’t know about. After 1987, funds switched to the manager-of-managers approach for more control. But now we’re finding a number of managers are saying, “You give us your objectives, and we’ll apply our knowledge and our work to those.”
The Alaska Permanent Fund has embraced this approach. When their new CIO joined he realised he couldn’t persuade great investment people to relocate to Juneau. So instead he got some smart investors at famous asset managers around the world and gave each of them a mandate. He could then have a good dialogue with each of them – in terms of how they’re meeting particular objectives – and he could use some of those ideas in the rest of the portfolio. I think you’re seeing a little bit more of that happening now, because no one manager is the source of all good ideas.
This way, managers can manage a $500 million mandate and get paid for what they’re doing — without impacting their own capacity. And funds can leverage the insights that they get from observing the way smart investors tackle the issues facing the fund.
LC: A lot of people think the first thing to internalise is passive. They think they can do it cheaper. I’m unconvinced by that. Funds should take in house those where they can add a lot of value. An index manager should be able to do passive cheaper than a fund when fully costed.
DH: That’s right. Cheaper and better probably, and you’ve got clear governance. If you’ve got systematic-type strategies, then there’s a question as to how easy it is to implement them internally, compared to telling an asset manager the formula you want them to apply. You work together to determine the formula and then it’s basically an execution function.
It doesn’t make a lot of sense to internalise passive management to save cost. All you’re doing is taking out one of the cheapest components of your investment management program, and you’re replacing it with, potentially, a lot of problems internally.
LC: What cultural impacts do you think there are with managing money inhouse?
DH: Most important is what happens when you get an active program that is underperforming. How does the business cope with that? How does the marketing team cope with underperformance because the internal team has underperformed for this particular year?
Unfortunately, one bad year’s numbers can affect your three- and five-year results. Then your marketing group, communications people and call centre are dealing with these issues. Because it’s the inhouse team, culturally, it can be quite damaging.
LC: Do you think the decision to manage money internally means you have to put more focus on what your investment beliefs are, and to be very clear about them?
DH: We have very clear investment philosophies and beliefs, which give a foundation to all of our investment decisions. But I think if you’re going to internally manage money, then you have to have another subsidiary set of beliefs that deal with that particular investment function. For example, why do you think that this particular function is going to be better done internally than externally?
LC: You mentioned the governance structure. How does the role of the investment committee change if you decide to internalise? Is there a different set of skills that are required?
DH: That’s another reason why you’d have a separate company. You’d still have the trustee board but you would have “Superfund Investors Proprietary Limited” or whatever as a separate company. Then it would have its own board. You’d probably have one or two directors from the main board involved, but you’d populate that board with people with fund management experience.
LC: Do you think it requires investment committee changes if you internalise asset management or internalise some functions?
DH: It doesn’t have to if you treat it like an external manager, and it has its own investment committee as a separate function. Then you’ve got two investment committees – one at the fund level and one at the internal investment manager’s level.
If you’re going to do it internally without that separate company, then yes, you need a completely different investment committee. You need people who have managed money and have experience in reviewing investment processes at that level, and understand how those investment processes fit together.
LC: Do you think bringing investment management in-house changes the way in which a fund should think about operational risk? For example, would it have an impact on the way they provision for that in their reserves?
DH: Yes, particularly in derivatives. If you’ve got internal management of derivatives, the accounting cost is not the same as effective exposure. If you’re putting hundreds of thousands of dollars on futures contracts, then pretty quickly you’ve got a couple of billion dollars worth of exposure. So you need to have proper pre-trade compliance and controls.
I also think the capital you need to have to cover operating losses goes up because there is an extra layer of potential risk there.
LC: So, in summary, what are the primary things to consider when deciding whether or not to insource?
DH: The only thing is net returns to members. Are you going to get better net returns for members with cheaper fees and similar performance? Even with slightly lower performance, you’re paying less, so the net performance can still be better.
Within the primary focus on net returns there’s also the ownership structure. For example, with infrastructure assets, we are increasingly only going into these deals where we can control our own exit.
The other thing you want is alignment: not only in terms of the ownership of the assets, but also the fee structures. Managers, under many pricing models, are encouraged to just gather assets, as opposed to getting performance for existing clients.
Net returns for members is the endgame, which can come from lower fees, lower costs or a better aligned ownership structure. Or, it can come from better alignment of the fund objectives and the objectives of the fund’s management function. It can come from more efficient tax management or reduced management costs. If you can meaningfully reduce any of these elements through internal management, to give better net returns to members, then it makes sense to consider it.
Internalisation is a bit of a trend that’s developing at the moment. But I’d emphasise that, having worked in investment management for more than a decade, I can see the strength of the manager-of-managers model.
I think ultimately that the manager-of-managers model is a long-term, stable structure. It’s just a matter of making sure you don’t get ripped off in the economics as you get bigger and bigger. Since the pricing models of a lot of investment management companies are not really designed around growing clients, superfunds are being forced to at least consider insourcing.
