By David Goodman
Funds can definitely consider internalising certain aspects of their investment management process. The large funds will not necessarily withdraw externalised functions in certain asset classes in total, they will look to internalise only those functions where it can be done efficiently. Before considering the detail of the impact of these changes it is necessary to understand that the drivers for change at a fund, as well as a fund’s view of efficient production, may differ to that held by other market participants.
The effect of scale in a market where there is a general model of management fees charged on the basis of NAV is well understood. For an index fund, there is an asset size above which the cost of internal production is cheaper than outsourcing. Passive equity fees are very low to reflect this and extremely large scale is needed to consider this move. When one considers the operational risk involved in passive management, outsourcing of this risk also has value to a fund. Funds can often consider internalising this type of activity when it is seen as a step in the direction of internalising other related mandates.
One of the most powerful drivers for change surrounds the experiences of various funds during the GFC. For many funds the greatest disturbance that occurred to them involved the failure of financial infrastructure that drove the structuring of many of the perceived lower risk (but larger value) assets in their portfolio. Redemption issues in index portfolios are still problematic for some funds due to frozen (now slowly thawing) collateral pools backing stock lending aimed at enhancing the returns of the fund. Some enhanced cash portfolios were involved in assets central to the GFC. Investors in long/short extension funds can now understand rehypothecation. For an asset owner these events may initially be a powerful driver to withdraw from these types of activities. Medium term it could be the case that this will cause funds to want to internalise the management of the pipes that froze during the GFC. Internally managing collateral pools and cash assets are related and now supportable by easily accessible technology and services. The GFC has created an incentive for asset owners to want to have more control over where their assets are.
Many large, traditional asset custodians to pension plans around the world have been actively seeking and winning mandates to provide back and middle operations support to large investment managers. These investment managers have adopted varied approaches and this has created the need for the custodians to segment the services they “in source” from some potential clients who are not looking for a full service solution. These services have great scale. An asset owner who has appointed a custodian to hold custody of their assets can now reach in and access the technologies that are provided to their investment to manage such things as collateral pools at a very effective price. This may enable asset owners to continue to enrol in securities lending programs (or ones they directly manage themselves) with a far higher degree of comfort and control over the operational risk they are running.
Broker supplied EMS technology has been abundantly available for a significant period of time. The above trends relating to investment managers outsourcing functions to custodians have provided services such as OMS and Portfolio accounting systems at an efficient price into which the EMS can be plugged.
So after establishing that it is now easier for funds to “in source” parts of their investment process and that they have powerful internal control reasons to do this, what will be “in sourced” and what are the implications for service providers?
As discussed above, aspects of listed equities could be internalised but the counter proposition exists that it maybe more efficient to externalise passive management and to internalise control of the operational risk surrounding any return enhancing activity undertaken. Assembling a team of high quality investment analysts capable of developing processes that can identify and capture sources of alpha in the long term is an expensive process. Using quantitative techniques to identify the betas that potentially drive these returns and internally implementing portfolios to create these returns is more likely. This would mean potentially lower allocations to managers whose returns are correlated to these systemic betas and higher returns to managers who can show uncorrelated alpha generation. Brokers will need to provide quant based product in exchange for implementation flow and structured product solutions that enables portfolio construction.
In addition to the ever cheapening technology required to support internalisation, an investment in human capital is required that would be sourced from traditional buy-side and sell-side firms. Money and geography are factors to be considered. Asset owners are more geographically spread than the large buy-side and sell-side firms and many have strong connections with their local communities and governments. Attracting experienced talent for these funds to internalise could be a significant barrier to entry.
Once an internal funds management team is created it can be employed to re- internalise aspects of other processes that may have been previously out sourced.
The management of trading costs is necessarily an important focus for investment managers as they seek to efficiently implement internally generated investment ideas and manage fund liquidity. Investment managers employed by asset owners see only their part of a client’s total fund. Investment managers are normally not aware of potential internal market liquidity that could exist in other asset pools managed on behalf of their clients fund and excess trading costs could accrue to the asset owner from “missed” internal crossing opportunities. An asset owner would normally have a better view of the cash flow profile of a fund, recognising times when it needs to raise or invest cash. In a fully externalised environment over trading could occur if cash flow moves in a different direction to cash needs. Some asset owners may take the view that they are better placed to manage the after tax performance of their fund centrally, than a number of external managers all trying to be internally tax efficient.
As a result of the above factors some funds which have built out internal capabilities are moving to models that allow some form of control over the underlying trading of external managers. Emulation funds, either as trackers or harvesters of optimised portfolios out of underlying manager sets, is one asset owner solution. Some part of the relationship between investment manager and client becomes Intellectual Propertybased rather than manufacturing based. Similarly another market based solution is for the creation of a centralised dealing desk within an asset owner where most fund orders are passed through on a daily basis. Emulation fund structures and centralised dealing desks also need to manage the payment of research based commissions to brokers who have provided input into the decision making process at their managers. Asset owners who adopt these types of processes need to be aware that potential implementation slippage can occur if the structure creates a timing delay between when the manager could have implemented and when the new structure does implement.
The growth of asset managers with internal trading desks and portfolio management capabilities is also creating great change in the transition management process. Much debate has existed about the merits of the competing buy side versus sell side provider models at a time when many of the largest clients are quietly moving on past this issue. We are seeing an evolution somewhat akin to the growth of portfolio trading and DMA usage at investment managers. Asset owner desks are working with technologies provided by traditional transition managers and others with risk management system providers to directly access the most efficient execution venues in order to implement underlying portfolio changes. For the transition manager the traditional transition client will continue to exist, in some markets purely for regulatory reasons. It is, however, likely that by value the largest asset owners will continue to move to a more internalised model.
As discussed above, one impact of the GFC has been a drive for some funds to hold assets closer to themselves, in order to manage more effectively the operational risk of some aspects of investing in financial markets. Prior to the GFC, but far from forgotten, was the performance of a number of quant related funds during 2007. Large asset owners have reacted to this by either investing in risk management models or utilising risk advisory services provided by parties such as custodians and transition managers. Having acquired the information and married with the internal trading capability, some funds are seeking to manage identified risks through the use of derivatives such as swaps that target reducing risks that are above a fund’s comfort level. The increased complexity and breadth of the ETF market globally is also allowing funds another efficient mechanism to manage risks within their portfolios. For many funds, this activity is just an extension of previously internalised processes such as asset rebalancing using very similar instruments.
A final and potentially most important implication of this trend for service providers, in particular brokers, is that a client class is re-emerging. Many years ago most pension savings were held in entities or invested in entities that operated as balanced fund managers. Services and research were provided to these clients that allowed them to make decisions across asset classes. Over a number of years industry best practice turned to one where asset owners chose to outsource their asset management to asset class specialists and managed asset allocation themselves or on the advice of another party. In this world the main financial relationship between funds and brokers was through specialised asset class managers and internal transition teams on an episodic basis. In order to service the asset managers, brokers themselves adopted similar specialisation. The internalisation process we are seeing now is in some ways a re-birth of parts of the old balanced managers. As these balanced managers will, by their nature, be making cross asset class decisions and, when implementing them, broker service providers will need to ensure their internal structures are in place to provide research input and execution services across asset classes.
By David Goodman