For us it comes from the cost and performance data we have collected from investment managers using the CTI (Cost Transparency Initiative) templates.
By way of background, the CTI standard is the best practice cost disclosure framework (or ‘template’) prepared for the UK FCA by its Institutional Disclosure Working Group. It is a guide for both asset owners and asset managers on how they should communicate the costs and performance of the products and services provided by managers. It covers all asset classes (both liquid and illiquid), applies to all fund types, and is suitable for any institutional asset owner. And it works.
Over the last two years we have collected and interpreted CTI data for almost 900 UK DB funds, collecting over 10,000 portfolios-worth of data from some 420 asset managers.
So what sort of expected things have we discovered buried in all this data?
Well, we now know the mean and median price points and net performance for over 32 different asset classes, where the scale economies lie, whether you actually do get more performance for paying more and so on. In other words, we know where Value (performance) for Money (cost) really exists. We also know a lot about the UK pensions landscape, its asset allocation, its preferred used of fund types and so on. All this is expected, but what of the ‘unexpected’ discoveries?
We recently ran a thought experiment through our benchmark database and came up with the median total cost and median net performance for a new client we were just starting to engage with – this was based on the proportional representation and assets under management (AUM) of each asset class as shown in the scheme’s Statement of Investment Performance. In other words, it was possible to calculate what the scheme should be paying and the net performance it should be achieving given the stated asset allocation. We were also able to calculate what the scheme could pay and the performance it could achieve on a best quartile basis.
More importantly we could go one stage further as the SIP not only listed the desired asset allocation, but also the managers and funds the scheme was actually using for each asset class. We could therefore calculate, with a high degree of certainty, what the scheme actually paid last year and how it performed, and all from the online SIP. Certainly, it made for an interesting meeting and it has whetted our appetite to see how accurate our various calculations are when the precise data has been gathered.
And to answer the question you might have, the difference between should and could for this scheme was approximately 15bps, or a 25% cost reduction, with an uplift in net performance of 2%. And the asset classes that had the most room for improvement on both a cost and performance basis were emerging markets passive equity and active corporate bonds.
Of course, cutting costs can be achieved in the main through negotiation if you have a basis for that negotiation (hence benchmarking). However, improving net performance might only be achieved by changing managers, which comes at a cost (transitioning).
The really exciting conclusion from all this is an attempt to calculate how much room there currently is in the DB world for cost reduction and performance uplift. Because if the answer is a 25% cost reduction across the board, that means an annual saving to the pension industry of approximately 16bps (the mean total cost of asset management of a UK scheme is 65bps, based on ClearGlass data). That’s £3 billion of cost savings each year.
We can’t wait to see if this is accurate.
Chris Sier, Executive Chairman, ClearGlass.
