This is the first in a series of articles covering some of the predominant issues regarding financial and economic statistics. This post seeks to introduce why and how statistics matter for investors.
The Why is simple: Investors require information to form expectations and make decisions; help them to comprehend how everything works; and convince stakeholders of their worth.
Investing is the art of calculated risk-taking. Blindly picking heads or tails is outright gambling. Statistics help to sway investors towards a belief in one outcome or another. Data permeates the execution stage of the institutional investment management process. The asset allocation stage requires portfolio managers to form return expectations for asset classes. Security selection requires a further refinement of views. Data is necessary for scenario analysis, back-testing, and stress-testing.
Statistics are combined with historical precedents to shape investor’s expectations – sometimes to their detriment. “While investors are often criticized for having too short a memory, it is also true that they keep fighting the previous war [emphasis added]” (Ilmanen, 2011, pp. 101).
Investment firms have to sell themselves to stakeholders and utilize data to this end.
How statistics matter is a function of their usefulness. Most investors would consider a statistic to be useful if it is relevant, representative, timely, exclusive, and comparable.
The type of stats called upon by investors is seemingly endless: real economic/company fundamentals, financial prices, and instrument-specific information, expectations, and alternative data.
The recent explosion in the usage and monetization of alternative data highlights the diverse ranges of data modern investors are incorporating into the investment process. Alternative datasets provide investment insights outside traditional company filings. The alternative data revolution is a product of the digital age.
Investors recognize that a single data point in isolation is out of context. Investors will expand on a data point in four separate dimensions to add perspective. Data is not useful or used in isolation, it must be amenable to analysis. Analysis is when we expand beyond a single data point.
The first data dimension is Time. Time is a key ingredient. Comparing the log of one data point with another defines the rate of change. Technical Analysis – an ancient practice – is built on the idea that patterns recur in time. The history of economic fundamentals is imprinted in investors’ minds making them susceptible to epochal forecasting errors in periods of violent structural change. Note the persistence of inflation expectations following the Volker disinflation.
Fundamental economic principles inspire the standard set of data used to characterize any real economic or financial market. This includes the flows of new production or financial market issuances; the product sales or demand; the change in inventory or total stock outstanding; and the price that clears the market.
The third data dimension is the spatial and relative value dimension: should I invest in Germany or the United States? Copper or oil? Stocks or bonds? This requires a database of comparable variables: economic concepts and valuations across investable nations, sectors, and asset classes. Relative values can be intra or inter-market.