New research finds institutional investors favour investing in private equity general partners that are located in the same state as their organisation. This is intuitive as local knowledge and contacts attract investments. However research from Stanford University finds this bias is coming at a large cost – about $1.2 billion a year – particularly for public pension funds.
Professor of finance at Stanford, Joshua Rauh, and co-author Yael Hochberg, are the first academics to produce a paper that examines the geographical bias in LP private equity investments.
“Institutional investor asset allocations to alternative asset classes have increased over the past few years but few papers explore institutional investors’ choice within alternative asset classes, and how that affects performance.”
Rauh’s research, outlined in the paper Local overweighting and underperformance: evidence from limited partner private equity investments, looks at the extent and cost of a particular strategy, showing a clear preference for home state private equity investments among institutional investors.
It finds, clearly, that limited partners (LPs) in a state overweight with general partners (GPs) in that same state.
This is true of all types of investors, including endowments and foundations and public and corporate pension funds, but is particularly true of public pension funds.
Rauh presented to delegates at the www.top1000funds.com investment think tank, made up of chief investment officers of public and corporate funds as well as foundations and endowments.
He showed that for all investors in the universe that he examined there was an 8.1 per cent excess overweight allocation to their home state.
This was determined by looking at how an investor, for example a California LP, invested in California relative to the entire population of US LP investors, and also compared to non-California LPs only.
While the overweighting was true for all types of investors, in public pension funds the overweight was more exaggerated with a 9.8 per cent home-state overweighting, he says.
In its own right this might not be a particularly unusual finding, but what is important for public pension fund investors is that the overweighting to home state also corresponds with underperformance.
The research looked at how home-state investments perform relative to out of state investments, examining whether they can make use of the local connections and networks and better access and information.
All funds measured had an IRR that was 2.86 per cent worse in their home state, but public funds had 3.74 per cent worse.
Endowments and foundations, the research finds, do not underperform in their in-state investments.
“Most hypotheses predict public pension funds overweight with a view to gaining excess positive performance on in-state investments because of detail on deals, superior information on in-state investments, rationed out of state access. But better information and better access is not translating to better performance,” Rauh says.
The political argument is that public funds are supporting businesses in their local communities and the chief investment officers from public funds echoed this view – that reasons other than performance often drove the investments in local investments.
Eric Baggessen, senior investment officer and head of asset allocation, CalPERS says how systems have “come in to being” is important because it means their governance will be different which in turn affects their investment allocations.
“For example in California given the government mechanisms underlying it, it is hard to separate it from the political environment. It would be like having negative gravity.”
He suggested Rauh add in governance considerations to the data or questions. But Rauh said the authors had added in board composition data but didn’t find anything.
Baggessen says it is impossible to take a state public fund and have it not affected by politics.
“They don’t exist in a political vacuum,” he says. “We have 13 board members; six are elected by unions and the other are from legislature appointees. It is hard to imagine they would get on board without a sense of political connection.”
Similarly Farouki Majeed, chief investment officer of the Ohio Public Schools Retirement System and former employee of CalPERS, says the results are not surprising.
“My experience in Minnesota, California and Ohio public funds is that there is board directed allocations to in-state funds,” he says. “Boards are not concerned if they underperform, what’s important is that money is allocated to the state, jobs are created, and enterprises are supported.”
Mary Cahill, chief investment officer of the Emory University endowment agreed.
“These investors are not saying we’ll live with underperformance but performance is not the main objective.”
Rauh conceded that home bias and underperformance may not be sub-optimal because these investments might support the local economy, and have positive externalities from local economic development.
This study did not measure the benefits of in-state investments, and Rauh said a future academic study could examine what are the benefits of in-state investments.
Rauh’s research looked at 19,000 LP/GP private equity fund investments by institutions over 1980 to 2009 drawing on data from Thomson Reuters’ Venture Economics, Preqin, VentureOne, and Capital IQ.
Broken down by type, 30 per cent of the investments were in buyout, 30 per cent in venture capital, 13 per cent in real estate, and 27 per cent in other (including distressed debt, mezzanine, fund of funds, secondaries).
“We looked at the share of all investments made in that state. For example Minnesota has 0.79 per cent of sample investments made in that state. So Minnesota LPs should invest 0.79 per cent in that state, like all other US funds. But they don’t, it’s 9.7 per cent of investments made by Minnesota LPs into Minnesota GPs. They are overweighting by a lot.
“This has implications for risk management,” Rauh says. “Risk management suggests that the optimal investment portfolio should underweight your own state. The state’s basket is already in the local economy, so they should diversify outside of that.”
The states LPs which are overweighting the most relative to the entire LP universe measured by Rauh are: Massachusetts (overweight by 32.4 per cent), Ohio (31.9 per cent), Illinois (24.3 per cent), Pennsylvania (16 per cent), and Minnesota (14.1 per cent).
Professor of history at Princeton and co-host of the roundtable, Stephen Kotkin, asked whether LPs were investing in their home state deliberately.
“The hypothesis is that you know people better and can judge them, so you would expect a correlation between locality and the result. But you’re showing the result isn’t there?” he asked.
“The positive result is more than missing. There’s actually underperformance,” Rauh says.
If each public pension LP performed as well as in-state as it does out-of-state, the LPs would reap $1.25 billion extra annually, Rauh estimates.
He uses Massachusetts, which is the state overweighting the most relative to the entire LP universe, as an example.
Massachusetts LPs investing in Massachusetts GPs had a net IRR of -7.5 per cent. Out of state LPs investing in Massachusetts GPs had an IRR of 1.93, and Massachusetts LPs investing out of state GP had an IRR of 1.4.
“This is an underperformance of around 9 per cent per year which when multiplied over the size of the $5.9 billion Massachusetts private equity program is a big cost,” he says.
The delegates also discussed why there is a difference between the performance of LPs out of state and LPs in state investing in a particular state.
BNY Mellon’s head of manager research and performance analytics Jamie Lewin, asked why LPs out of state doing a better job of choosing GPs in state.
Majeed posited that it was adverse selection and because there are allocations across the state, there are more investments so more managers (including the bad ones).
Other issues for discussion were raised by the chief investment officer delegates including the staff model and delegation within public pension funds (Stan Rupnik, Illinois Teachers’ Retirement System) and the role of size with regard to investments and GP capacity (David Cooper, Indiana Public Retirement System).
Breakdown of type of LP – overweight to home-state GPs
Public pension – mean excess in state 16.5 per cent
Private sector 7.8 per cent
Endowment 8.1 per cent
Foundation 9.6 per cent