Rethinking corporate finance theory

New research finds the practice of benchmarking in the asset management industry affects corporate investment and other decisions

By Anna Pavlova 17 January 2019

Benchmar-inclusion-banner

The use of benchmarks to measure both stock-price movements and the performance of asset managers in relation to market averages is so widespread as to be taken for granted. Assets under management total $85 trillion globally, a figure that is expected to grow further, and most of those funds are managed against benchmarks.

One particularly striking figure comes from S&P Global, the financial intelligence group, which reports that, at the end of 2017, there was just under $10 trillion managed against the Standard & Poor’s 500 index alone. In the AM industry, performance is normally evaluated relative to benchmarks. Otherwise, how would one compare a bond fund manager with a stock fund manager? The two asset classes have very different risk. The managers’ benchmark-adjusted performance, however, is more of an apples-to-apples comparison. This use of benchmarks, in turn, creates incentives for managers to allocate a fraction of their assets under management into benchmark stocks. They do so irrespective of characteristics of benchmark stocks, and in particular regardless of their risk.

There has been some research into the effect of benchmarking on the price of stocks included in benchmark indices as opposed to those excluded, and into the differing impacts of benchmarking on, respectively, institutional and retail investors. But, until now, there has been no analysis of the effect of benchmarking on corporate decision-making.

The benchmark inclusion subsidy

Together with Anil K Kashyap, Professor of Economics and Finance at the University of Chicago Booth School of Business, Natalia Kovrijnykh, Associate Professor of Economics at Arizona State University and Jian Li, PhD candidate in economics at the University of Chicago, I have sought to remedy this deficiency, but in doing so have found ourselves cutting across conventional corporate finance theory, which holds that the value of an investment project is calculated solely in relation to its cash-flow risk.

One of the main things we stress when we teach basic corporate finance is that you should not confuse the cost of funding for a firm’s assets in place with the appropriate discount rate for new projects. The most extreme version of this is to explain why public utilities do not have a comparative advantage in funding fintech start ups. They might have some debt capacity that is idle, but that does not mean they should apply that to an unrelated project with different risk. We argue that benchmarking creates an exception to this rule.

Holding the benchmark is a zero-risk proposition for a fund manager. Diverging from the benchmark is risky.

Our findings are that the need for asset managers to hold benchmark stocks gives the companies that issue such stocks a different cost of capital from issuers of non-benchmark stocks. Asset managers have to hold those stocks regardless of the issuing companies’ cash-flow risks, thus the companies concerned can adopt different criteria with regard to mergers and acquisitions, spin-offs and so forth.

Investing in a project will increase a benchmark firm’s share value by more than it would that of a non-benchmark firm. This, we argue, is the observable phenomenon that we have called the “benchmark inclusion subsidy”. We completely characterize the nature of the subsidy. This allows us to predict the types of firms for which its presence will be most important. Prior research has at best thought about average index effects (or benchmark effects). We can go a step further.

Our title may raise hackles straight away, given that “subsidy” is a word more usually associated with government handouts than with the workings of the asset management industry. My colleagues and I believe, however, that it is justified, in that the behaviour of asset managers in relation to benchmark stocks amounts to a subsidy to the entities concerned.

To see why, let’s first look at the effect of benchmarking on asset-manager behaviour. Were you to ask such a manager for their attitude to benchmark stocks, they might well reply: “If I don’t have a view on the stock I will buy it in the exact weight as in the benchmark. If I have a positive view I shall buy and slightly overweight it. If I have a negative view, I will still buy, but underweight it.”

Put bluntly: Holding the benchmark is a zero-risk proposition for a fund manager. Diverging from the benchmark is risky. In light of this, it is perhaps unsurprising that, on average, a company’s stock price will rise by six per cent on its inclusion in a benchmark.

Two types of investors, two kinds of behaviour


Now compare the attitudes of asset managers and conventional investors. Asset managers have a different level of demand for stocks in the benchmark and for those outside it. Conventional investors do not.

A conventional investor is seeking a reward from their investment, one calibrated by measuring the level of risk they deem acceptable against the return they seek, the “mean variance portfolio”. Benchmarks have no real significance for them.

Asset managers, too, will invest in the mean variance portfolio. But they will also hold the stocks in the benchmark portfolio. Their pay is tied to performance which, in turn, is measured against benchmarks. To illustrate the different behaviour of these two types of investors, imagine two firms, firm 1 and firm 2, which want to acquire firm Y. Assume firm 1 is in the benchmark and firm 2 is outside. Assume, for now, that the firms’ cash flows are uncorrelated. In a world without asset managers, demand for firm 1 would be unaffected by its inclusion in the benchmark.

But once asset managers enter the picture, there will be an automatic demand for asset 1 over and above the level of demand that would be generated by conventional investors.

Now, let’s discuss a merger of firm 2, which is outside the benchmark, with firm Y. Asset managers do not have any specific interest to hold firm 2, because it is outside the benchmark. The merged entity would then be priced at simply the sum of the two parts. There would be no value created from this merger.

Once firm 1, which is in the benchmark, merges with firm Y and thereby brings firm Y in the benchmark, there is a big difference. The mechanical demand from asset manager for firm 1 extends to any assets that this firm owns, including the newly acquired form Y. The combined entity turns out to be more valuable than the sum of its parts, because of the inclusion in the merger of asset 1, a benchmark stock. This is the elusive financial synergy, which many corporate finance textbooks say does not exist.

Interestingly, the riskier firm Y is, the more it gains from the merger. Remember, once it has made it to the benchmark, asset managers are willing to invest in this firm regardless of its riskiness. Similarly, the more funds under management are following the benchmark, the bigger the subsidy. So a benchmark firm would be willing to undertake a merger that a firm outside the benchmark would not.

Beyond the benchmark

Now let’s introduce a correlation among firms’ cash flows. Intriguingly, the benchmark inclusion subsidy does not vanish once we step outside the benchmark. That over-priced benchmark stock is less attractive to ordinary investors. So, instead they may well buy a stock that is outside the benchmark but which has many of the characteristics with the benchmark stock but is cheaper. In equilibrium, of course, this implies that the non-benchmark stock will also be overpriced, but not by as much.

As we can see, there is a continuum in the subsidy beyond the benchmark. The subsidy does not fall off a cliff edge just because a share is outside the benchmark, but its operation is less pronounced. Projects or assets that are more correlated with the benchmark receive a bigger subsidy.

no investment professional wants to be told that they are systematically holding over-priced benchmark stocks

By contrast, there is no subsidy for risk-free projects.

Finally, the subsidy grows in size in parallel with the increase in assets under management that are following that particular benchmark. It increases also when there is a rise in the number of passive funds tracking that benchmark.

In some ways, it is remarkable that the benchmark inclusion subsidy has escaped scrutiny for so long. In part, I suspect, this is because, as mentioned earlier, the whole notion flatly contradicts the conventional theory of how firms are supposed to arrive at decisions covering investments, mergers and other key issues. This theory is “entity agnostic”, assuming that cash-flow risk is assessed identically regardless of whether the firm’s stock is in a benchmark or not.

I hope we have shown this is not the case, and that benchmark firms will take on projects that non-benchmark stocks would not.

In part, also, it may be because the implications for asset managers are uncomfortable, given that no investment professional wants to be told that they are systematically holding over-priced benchmark stocks and potentially distorting funding for different types of firms.

But ignoring the benchmark inclusion subsidy will not make it go away. The asset management sector is big and it is there to stay. As it continues to grow, the distortion will only get bigger.

Comments 3

Evgeny 4 months and 12 days ago

All companies do the same things, they coping each others in strategies, they hire the same consultancy firm, they implement TQM, ISO, 6Sigma, benchmark and so on. All that leads to strategic convergence and many companies simply disappear. In that sence benchmark is not the best idea

Evgeny 4 months and 12 days ago

All companies do the same things, they coping each others in strategies, they hire the same consultancy firm, they implement TQM, ISO, 6Sigma, benchmark and so on. All that leads to strategic convergence and many companies simply disappear. In that sence benchmark is not the best idea

Blue-eyes 5 months and 1 days ago

There is still risk. The risk of the firm leaving the benchmark. Consequently firms more at risk from leaving (not necessarily because of bad performance - only bad relative performance is sufficient) should have a lower premium than those less at risk.