Fact File: Calculating the Illusive Size of “Mr. Market”
When it comes to explaining the equity market’s mysterious day-to-day moves, the financial media love nothing more than a convenient explanation. “Today ‘the market’ was up . . . ” or “Investors ran for the exits today . . . ” are common refrains. But what or who, exactly, is “The Market”? Which investors? What is implied, but never fully explored, is that there somehow exists a consensus among all investors about how a given day’s news is discounted.
Can one day’s worth of trading activity be legitimately generalized to represent the opinion of all investors? Using turnover ratio data provided by the World Bank for key domestic equity markets, along with the knowledge that there are, on average, 251.6165 trading days in a given year, I analyzed two decades worth of data to arrive at a time-series showing the average daily number of traded shares outstanding in eight markets.
Sources: CFA Institute, World Bank. Note: Turnover data must be uncompounded to back out the average daily number of traded shares outstanding.
So what can we learn?
For starters, the average percentage of the market trading daily over the last twenty years reached a crescendo in the United States in 2008 — at just 0.64%. Yet this represents a turnover ratio of a whopping 404.07%. The year 2008 also marked the global peak of trading activity, with 0.35% of all shares outstanding in all eight markets trading each day.
In each year subsequent to the peak of 2008, the number of shares outstanding that are trading has dropped each year. Even our friends in the financial media would agree that a peak figure of shares trading per day of 0.64% is surprisingly low. After all, when we listen to the day’s news and trading activity is presented as “The Market,” the perception is of a large consensus of investors all discounting news in the same way — not a mere fraction of less than 1% of investors.
Also of interest in the data is the gradual buildup of trading froth during the dot-com era, the subsequent decline, and then a new buildup during the real estate bubble, which, as we all well know, also declined. This suggests that one way of tracking whether or not markets are in the midst of the bubble is to see if the average percentage of the market that trades daily is rising.
Please note that the content of this site should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute.
6 thoughts on “Fact File: Calculating the Illusive Size of “Mr. Market””
It would be interesting to see what the table would look like if you controlled for buy-and-hold investors.
Estimate the percentage of total float is owned by entities who have an average turnover of 25% or less (or a widely held threshold for consideration as low-turnover), then recalculate the table.
Thank you for your comment and suggestion.
Unfortunately the World Bank does not provide granularity within the statistics that are publicly available so I cannot execute the research you suggest, above. I agree it would be interesting to see what the percentages would look like. I would love to create multiple thresholds, say 10% turnover, up through 400% turnover to see how it all shakes out.
A very interesting analysis. However, referring to the last statement, I would also take into consideration the daily trading volumes of F&O’s, while trying to analyse whether we are ‘in the midst of a bubble’ or not.
Again, very nice Jason.
And thanks for the link here.
As you mention above, surprisingly low numbers.
Perhaps a means of taking into account buy-and-hold would be to estimate the average free float within each market by recreating the respective major indices at specific time points. Working backwards from your quoted values, you should be able to find some fairly reasonable range for non-B&H trading activity.
I very much agree the premise for the research. The financial press, who when all said and done are simply creating the journalism that their audience demand, are fixated for finding a reason in the randomness of the market. CNBC and CNN would not get through one episode if you banned them making some reference to what “The market” wants to do.
It is not going to happen, but it would show some refreshing insight to hear:
“The market was up today for no particular reason”
“The S&P 500 was down today because more was for sale than was demanded at the price”
Fixating on reasons for market moves which quite possibly don’t exist (or if they do are not those which we can understand and analyse) is somewhere between pointless, dangerous, and a self fulfilling prophecy.
Can you please give more detail as to how 0.64% daily volume equates to 404.7% turnover ration. What is the calculation.
Also I do not see this from the data (from looking at the 2008 row). What am I misunderstanding?
“The year 2008 also marked the global peak of trading activity, with 0.35% of all shares outstanding in all eight markets trading each day.” I
Hi Nathan E.,
Thank you for your comments, it is nice to know that there is accord with my thinking out in the information sea.
How is the information calculated?
1. The World Bank provides its data as total annual turnover for a nation’s stock markets. So, for example, in the U.S. we have the 404.7% turnover of 2008.
2. Next, I have assumed that, just as returns are assumed to be random when we calculate compound annual returns, volumes that generate those returns must also be random. This assumption results in adjusting the turnover data for Brownian Motion. In finance terms that is simply the familiar ^1/n adjustment made to a compounded number to account for random movement of returns through time.
3. Given the above assumption we have to discount the annual 404.7% turnover back to a daily turnover figure. The calculation is:
4. [ (1 + annual turnover) ^ (1 / average number of trading days in a year) ] – 1 = daily turnover; [ (1 + 4.047) ^ (1 / 251.6165) ] – 1 = 0.006454316 = 0.64%
As for the 0.35% number, that is simply the average of the 8 markets featured above. So it is the average across markets and not an intermarket figure. In other words, it is not based on total global trading volume divided by total global shares outstanding. I do not have data about absolute volumes or shares in individual markets. Taking the straight average across markets can be justified though because I am trying to demonstrate the size of “Mr. Market” and trading volume magnitudes in South Africa are probably highly correlated with others. If I had calculated based on intermarket activity it would give undue weight to the larger markets.
Hope that helps in your understanding.
PS – Thanks to all others that have made suggestions. As CFA Institute is a non-profit my access to data and other resources is largely restricted to those publicly available. So I cannot follow the lines of inquiry you have all graciously suggested.