Short-termism, reporting frequency and performance measurement
Short-termism is often attributed to quarterly reporting. The evidence on reporting frequency is mixed; the evidence on how returns are measured and how managers are rewarded is stronger. A simulation shows which projects are approved under different evaluation horizons.
A recurring argument in management writing is that quarterly earnings reports make companies short-sighted: managers cut long-term investment to hit the next number. The argument identifies a clear cause, the reporting calendar, and a simple remedy, less frequent reporting.
In the performance measurement module of my MBA, the course examined the case for and against. My conclusion was that reporting frequency is a minor factor. The main cause is the yardstick, what is measured and over what horizon, and from the incentives tied to it.
This article reviews the evidence, frames short-termism as a measurement problem, and presents a small simulation in which the same twenty projects are evaluated with different yardsticks.
The relevant literature is as follows.
Stein (1989) showed theoretically that managers can behave myopically even when markets are efficient: if investors infer value from current earnings, managers are tempted to boost earnings at the expense of long-term value, and rational investors expect them to.
Graham, Harvey and Rajgopal (2005) surveyed over 400 financial executives. A large majority said they would give up economic value, for example by delaying a positive-value project, to meet an earnings target or smooth earnings.
On reporting frequency the evidence is mixed. Kraft, Vashishtha and Venkatachalam (2018) linked the historical move to more frequent reporting in the US with lower investment. But the cleanest natural experiment, the UK's requirement of quarterly interim management statements from 2007 and its removal in 2014, showed no meaningful change in investment when companies stopped, as Rajgopal (2018) summarises.
On incentives, Mizik and Jacobson (2007) found that firms which appeared to inflate earnings around a share issue, for example by cutting marketing or R&D, were rewarded by the market at first and penalised substantially in later years.
On the yardstick, Martin (2020) argues that companies judge investments by return on book capital, while investors judge them against current market value. Kaplan and Norton (1992) proposed the balanced scorecard to add leading, non-financial measures, and Ittner and Larcker (2003) found that only 23% of the companies they studied had checked that those measures actually predicted results.
Evidence on reporting frequency
If quarterly reporting were the main cause of short-termism, removing it should have changed behaviour. In the UK it did not. Some US evidence points the other way, so the evidence is mixed and gives weak support for attributing short-termism to reporting frequency.
Evidence on performance measurement
Two features of ordinary accounting bias decisions systematically:
- Horizon. A manager judged on profit over the next two or three years sees costs now and benefits later. Any project whose payback falls outside the evaluation window looks like a loss.
- Expensing. Spending on research, training and security is mostly expensed rather than capitalised. It hits this year's profit in full and appears nowhere as an asset. When budgets get tight, these lines are the easiest to cut: the cut flatters the numbers immediately and the cost arrives years later. This pattern is familiar from security budgeting. Households know it as well: postponing the service of a car or the repair of a roof improves this month's budget and raises the cost of the year in which the repair can no longer wait.
Short-termism as a measurement problem
In engineering terms, this is a measurement problem: the sensor reports a quantity other than the one to be controlled, and optimising the reading moves the system away from the target. It is the same mechanism as the incentive distortions in performance measures derived from strategy, and the remedies follow from the framing:
- Protect strategic spending in a budget that is not traded off against quarterly targets.
- Judge at the margin and at current value, not against historical book capital.
- Pay over the consequence horizon. A decision that plays out over five years should be rewarded over five years.
- Use leading indicators, but validate them. An unvalidated indicator is another number that can be gamed.
The figure below varies the manager's yardstick. Judging investments by short-horizon accounting profit rejects projects that are clearly worth doing; lengthening the horizon recovers most of the value; and capitalising the cost helps only a little.
Twenty stylised projects, each with a cost now and a benefit every year for eight years after a delay of up to three years. The bars show each project's true value (net present value at 8%). A manager approves a project only if it improves the accounting profit they are judged on within their horizon.
Suggested settings:
- Three-year yardstick. With the defaults, only 3 of 20 projects are approved, capturing about 28% of the available value. Most bars above zero, representing projects worth doing, are grey, indicating rejection.
- Capitalisation. Turn on capitalisation, spreading each cost over five years. One more project gets through, and captured value rises to about 38%. Capitalisation helps, but the horizon remains the dominant factor.
- Five years. Turn capitalisation off and set the horizon to 5 years. About half the projects are approved and captured value jumps to around 77%.
- Very long horizon. Set the horizon to 12 years. Every project is approved, including one or two with negative value. An undiscounted long horizon has the opposite bias: money in year eleven counts as much as money today. A longer horizon alone therefore does not solve the problem. The yardstick has to measure discounted value.
Effect of a short horizon. A project with cost at and benefit in years has net present value
but a manager judged on undiscounted profit over years sees
When , for every project, however valuable: the yardstick cannot recognise its value. As grows, eventually exceeds the NPV, because it ignores discounting, so marginal projects with negative NPV start to pass. Only a measure that values cash flows at the right time ranks projects the way their value does.
None of this requires changing how often a company reports. It requires measuring the right thing over the right horizon and rewarding people for the outcome rather than the reading. For security leaders, it also explains why security budgets are so easy to cut and so hard to defend: they are expensed, their benefits are deferred and probabilistic, and the yardstick sees only the cost.
References
- J. C. Stein, "Efficient capital markets, inefficient firms: A model of myopic corporate behavior", Quarterly Journal of Economics, 104(4), 655-669, 1989.
- J. R. Graham, C. R. Harvey and S. Rajgopal, "The economic implications of corporate financial reporting", Journal of Accounting and Economics, 40(1-3), 3-73, 2005.
- A. G. Kraft, R. Vashishtha and M. Venkatachalam, "Frequent financial reporting and managerial myopia", The Accounting Review, 93(2), 249-275, 2018.
- S. Rajgopal, "What would happen if the U.S. stopped requiring quarterly earnings reports?", Harvard Business Review, 6 September 2018.
- N. Mizik and R. Jacobson, "The cost of myopic management", Harvard Business Review, 85(7/8), 2007.
- R. L. Martin, "What managers get wrong about capital", Harvard Business Review, 98(3), 2020.
- R. S. Kaplan and D. P. Norton, "The balanced scorecard: Measures that drive performance", Harvard Business Review, 70(1), 71-79, 1992.
- C. D. Ittner and D. F. Larcker, "Coming up short on nonfinancial performance measurement", Harvard Business Review, 81(11), 88-95, 2003.