John Eliopoulos, Investment Manager at Telstra Super talks through the thought process behind, and the journey of, internalising assets as a Superannuation Fund.
I think there are three primary motivating factors and a couple of other more subtle ones. The amount of regulatory change and complexity of the Australian Superannuation system has meant that administration and management costs have been driven higher. Like most organisations Telstar Super has a cost focus. We benchmark our costs against other Super funds and aspire to achieve bottom quartile expense ratios. If we can identify new ways to save our members money then this flows back as improved net returns. External management fees are calculated as a percentage of funds managed. However, funds managed internally have relatively small fixed costs which are a fraction of the typical fees charged by external managers. Therefore, internalisation is an efficient form of cost scalability.
The other two motivating factors are capacity constraints and availability. Sometimes successful external managers run out of room to take on any additional funds without compromising their ability to generate alpha. Internalisation can provide another source of capacity for a particular style that is capacity constrained in the open market. We have typically experienced this for the less liquid part of the market such as small cap stocks.
The third primary motivator is availability. Occasionally there may be a requirement that is unique to a Super Fund’s needs. This requirement may not be able to be satisfied from the available products/services currently in the market, either because they don’t exist or lack a track record of success. We have created several internal solutions where we believed it was better to build it rather than try and buy a compromise solution.
There are a couple of other motiving factors that are a little bit more subtle. The first is this concept of evolution. I believe that Super Funds start life as passive vessels of member retirement savings governed by a board of trustees. Through organic growth or consolidation they begin to enjoy economies of scale that facilitates the progression to a more self-contained institutional model. In-sourcing of previous tasks such as legal, accounting, operations, financial planning, and investment are some examples of this evolutionary internalisation process. At Telstra Super we are on that journey, somewhere on the spectrum, as we evolve towards a self-contained entity.
“We have Australian equities as one of the larger assets that we manage in-house, pretty much for the reason that it’s in our own backyard, it’s in our own time zone and we have a level of intimacy and engagement with that particular asset class that makes it natural for us to want to manage it in-house.”
The last motivating factor for internalisation is the level of market engagement that you can be exposed to from running internal portfolios. We have found that this has assisted in attracting and retaining key staff. It has also provided us with key insights on how our external fund managers are thinking about the portfolios they are managing on our behalf.
What is the decision process for what assets are internalised?
Our process in deciding which asset classes we choose to internalise is very similar to some other funds that have taken this path. This is best explained by outlining our thought process for each of the assets we have in-sourced. It’s worth mentioning that our current model of investment management is a hybrid, where we continue to use external managers in all asset classes even where we have in-sourced for particular asset classes.
We have Australian equities as one of the larger assets that we manage in-house, pretty much for the reason that it’s in our own backyard, it’s in our own time zone and we have a level of intimacy and engagement with that particular asset class that makes it natural for us to want to manage it in-house. We do property as well, because we prefer to actively negotiate the terms of our investment rather than passively accept what is being packaged up and sold. We manage our asset allocation overlay and rebalancing through the use of derivative instruments that we run internally. We believe it’s one of our core competencies and an integral part of managing our member’s funds. We also manage fixed interest for the same reasons mentioned earlier. Lastly we manage cash and foreign exchange, partially because these are relatively operational in nature as well as being comparatively less complex than the other asset classes.
Essentially our decision to internalise has to do with familiarity, operational complexity, and our desire to source negotiated proprietary deal flow. I believe it’s all part of this evolutionary journey that asset owners embark on as they get bigger, as they get more comfortable with the will to investment.
Is there a dollar figure at which point funds look at internalisation?
Identifying a notional dollar value as the key metric to trigger internalisation is probably the most visible and objective way to think about this question. However, it’s just one factor. I think you do need to have a critical mass of Assets Under Management (AUM) before you would even think about on this journey. Somewhere between $5billion – $10 billion AUM is a reasonable range to start in-sourcing some parts of the value chain. We were on this journey before $10 billion AUM; we’re at almost $15 billion at the moment. There’s another group here in Melbourne, Australia that has a large internal team that does pretty much what we do and they have a significantly smaller amount of AUM than us and have been doing it successfully for quite some time.
To read more buy-side content, click here.
The AUM figure isn’t necessarily the catalyst, but it’s certainly one consideration if as a firm you are considering internalisation because you do need to have a certain critical mass. It’s hard to put my finger on other considerations but sometimes legacy and history can play a part in whether or not you are going to internalise sooner rather than later or at all. There are certainly funds that are larger than us and they haven’t decided to internalise just yet.
From a technological standpoint, is it relatively easy and quick to get the assets up and running?
Everybody is different with regards to their requirements. There is no one solution that’s going to satisfy all the different Funds. It’s an incredibly bespoke outcome when you go down this road. Certainly what we’ve found is that it all sounds fine during the business planning stage, but until you actually start getting into operationalising the funds and experiencing the day to day trading and the day to day portfolio management, it’s then that you really begin to realise that the task is much bigger, much more complicated, and has a lot more moving parts to it then you would have envisaged when you first began the journey.
It is a very big consideration which can suck in a lot of resources. We are going through our third iteration at the moment of our whole middle office, our entire technology requirements and how everything links to the various components. We are putting pressure on our key service providers to help us along the journey, so we are still evolving and still learning about how to do things better.
Is it a growing trend industry-wide?
It appears to be at the moment. The desire to save on costs and “do it yourself” has even extended down to the retail level where we have seen the proliferation of self-managed Super Funds by individuals. Talking to my peers in the industry who do internalise or who are looking to, the first thing and the obvious thing is there’s got to be a pre-disposition at the board of trustees that it’s something that is desirable for the members.
Over the course of the journey our process can best be described as logical incrementalism, where we’ve taken small chunks of relatively simpler things and brought them in-house. It’s been done in small bite-sized pieces that can be managed in the event that something goes wrong. Once the level of confidence rises then we have gradually taken the next step. You bring a little bit more in and or you move on to another asset class or something else that adds to the overall effort of internalisation. That’s been the approach that others in the market appear to be taking too. They’re starting with the equities and cash and bringing in only a small percentage of the total AUM and going from there.
How much is enough in terms of internalisation?
The issue of how a fund derives its beta and alpha partially answer this question. If you have a philosophy that says, “We’re happy to just have beta in the underlying asset class and we will let our asset allocation generate all the alpha,” then it could be a quicker road to bring a large amount of money in-house. This is because adopting beta strategies like indexation is relatively simpler than an active management approach which requires a stock selection process and portfolio construction disciplines.
At Telstra Super we’re quite active. Everything we do is all about active management. We have a very strong philosophy that the markets are inefficient, and through skill and experience, we can exploit those inefficiencies. As a consequence, we presently have around one quarter of our Australian equities managed internally. We think that this is around the right proportion given our size and philosophy. The proportion varies from asset class to asset class depending on complexity and familiarity. For instance we currently have greater than 50% of our cash managed internally but no international equities. Overall around 20% of total assets are managed internally with a bias to move a little higher.
To summarise our active management philosophy has capped our ability to internalise too much, while our size has also required a certain amount to be internalised in order to “move the needle” on costs.