Quarterly win: Signal versus Noise, Cost vs. Benefit – CFA Institute Enterprising Investor

Because the White House trivializes the value of quarterly reporting for companies, investors are confronted with a well -known question: do the costs for producing information heavier than the benefits?

With the help of Robert Shiller’s long -term data, this message shows that quarterly income contains information that is probably valuable for both long -term allocators and short -term traders. The benefits that I do not try to quantify must be weighed against savings of less frequent reporting.

Quarter versus semi-annual: what is at stake

The White House this week asked for a change From every three -month to half -yearly winning report. President Donald Trump argued that such a shift would save companies money and time.

That can be true. But would investors lose valuable information?

To answer this question, I use profit data from Robert Shiller’s Online Data From January 1970 (1970: 1), the year in which the Securities and Exchange Commission made every three -month income mandatory until 2025: 6 to test relationships between the change in the three -month profit, a profit of six months and the trend in profit. I define the trend as a centered advanced average change in the profit of 61 months. In particular, I test whether knowing the changes of three months of income helps an investor better estimate the changes in the longer -term trend in the profit.

Graph 1 shows an income of three months in green profit of six months in red and trend income in blue. Series start in January 2000 (2000: 1), instead of 1970: 1, for the convenience of visualization.

Graph 1. 3 months, 6 months and trend income, 2000: 1 to 2025: 6.

Source: Robert Shiller Online data, calculations from the author.

Of course the income of three months choppier is then six months of income. But it is not clear from the visual inspection that knowing income of three months in addition to the six -month profit would help a long -term investor to predict changes in trend income. (I test this below and think they can).

However, it is clear that a short -term investor, who is perhaps interested in profit changes in periods of less than a year, would benefit from knowing three months of income. This observation is attached empirically below.

I start with the long -term investor, which I was taking on, is interested in the long -term trend in income. A natural way to measure the value of having three months of income in addition to (or instead of) six months of income, is modeling the change in trend income as a function of one or both, which estimates that model with ordinary smallest quadrates and the comparison of model accuracy. In this message I use R-Kwadraat as my benchmark for fit (or adapted R-Kwadraat)-how bigger, the better.

At any time the investor knows half of the current trend in the income. That is, they know the first 30 months of income from the current 61-month window, my proxy for the trend in income. And they know the last three months of income, or the last six months of income, or both.

To determine whether receiving income information every three months, unlike every six months, would help the long-term investor to better predict the trend, I estimate specifications in which the change in 30-month trend inflation is explained by the change in six-month profit only the earlier profit-trend change (model 1). In Model 2, trend change is explained by the same variables plus the three -month change in income. Results are displayed in Table 1.

Table 1. Regressions of trend inflation change on 3- and 6-Mongades of profit changes, 1970: 1- 2025: 6.

Source: Robert Shiller Online data, calculations from the author.

Because I am not interested in conclusion, I leave the discussion about estimated coefficient values, other than to note that they enter the expected sign. Notwithstanding this, I include the earlier trend in the profit to reduce bias in my estimates and standard errors appear in brackets next to any estimate.

The most important result is that the addition of quarterly profit (three months change) improves the fit-the custom R-squared rises of 0.098 for model 1 to 0.126 for Model 2. Although neither of them is impressive, these results suggest that quarterly profits can help predict the long-term profit. Other fit measures, namely the Akaike and Bayesian information criteria (AIC and BIC), confirm that the specification with income of 3 months is more accurate.

As for what can be interesting for traders (short -term investors), it could be guessed that the three -month profit change is related to the next change of three months. The quarterly changes of the quarter are indeed persistent. The spread in graph 2 shows the autocorrelation of three -month income, whereby extreme values ​​(profit changes larger than 100%) are removed for simpler viewing. The estimated slope is 0.601 (SE = 0.031) -The blue best fit is flatter than the black diagonal line of 45 degrees and the R-squared is 0.361.

Graph 2. Three months left behind profit change versus three months of profit change, 1970: 1-2025: 6.

Source: Robert Shiller Online data, calculations from the author.

And with the risk of estimating it obvious, the R-Kwadraat of a model that explains 12 months of income with a profit of six months (from six months earlier) 0.699, while recording income of three months (from three months before) improves the fit to 0.953.

Costs versus benefit

It is almost axiomatic that in most applications more data is preferred over less. And the results discussed here suggest that quarterly income contains valuable information for investors. But producing income is expensive.

As supervisors are considering reducing the reporting frequency, they must not only weigh the savings, but also the potential losses – losses for investors resulting from less transparency and to the economy as a result of reduced market efficiency.

More to think about

Past CFA Institute member surveys show clear support for quarterly profit.