Lies, D----ed Lies and Statistics

October 01, 2020

In a year when statistics are flying around like an infielder’s drill in baseball, we thought it prudent to touch on the confusion often present in common statistics.

While often mistakenly attributed to Mark Twain, one can find the explanation in Wikipedia that “"Lies, damned lies, and statistics" is a phrase describing the persuasive power of numbers, particularly the use of statistics to bolster weak arguments. It is also sometimes colloquially used to doubt statistics used to prove an opponent's point”. In this article, we will focus on the hospital Intensive Care Unit (ICU) capacity riddle, stock market gyrations, the Covid-19 infection rate, and short-term investment performance. Another article in this news- letter will focus on the oft confusing Unemployment Rate.

ICU Capacity. In mid-July, a colleague and I were puzzled how Miami-Dade County hospitals were operating at “107% ICU capacity”. By late July, it was announced that the same county was operating at “146% ICU capacity”. I thought I could solve the problem by digging into the CDC’s website with a myriad of statewide data, including total ICU rooms, Covid-occupied rooms, and the resulting utilization. It wasn’t until I asked my mother, a retired nursing administrator, who explained that each hospital has a specific number of “licensed ICU rooms”, ones that are appropriately sized and equipped, merely needing staff to become functional. Other patient rooms, if large enough, could be filled with addition- al machinery and staff, converting them into temporary ICU rooms. Many communities have set up M.A.S.H.-style tents or converted convention space (New York and Los Angeles borrowed Navy hospital ships for a short time) for added non-ICU room capacity, allowing hospitals to provide additional Covid-19 capacity. The combined total of these non-licensed ICU rooms explains how certain counties can operate above their licensed ICU capacity. In addition, hospitals were ordered to minimize elective surgeries and operate below half capacity (they typically operate at 90-100% of capacity), creating more expansion capability for Covid victims. If you dig hard enough for more information, the truth often reveals itself.

Stock Markets Snap Back. Earlier this year, we experienced the “fast- est correction in market history”, as the US. stock market dropped by over 10% in six days, entering bear market territory (a correction of over 20%) a few weeks later. In roughly one month, the S&P 500 index had lost 34% of its value “from peak to trough”. From this low point, the S&P 500 rose over 44% by June 7th. However, it would not reach the level of the previous high from mid-February until August 11th – how is this possible? This is merely a case of changing bases or denominators, as a bigger percentage increase from a lower value (51% in this case) is needed to fill in the 34% “hole” created by the market downturn.

Infection Rate. It may already be the most studied disease in human- kind, but Covid-19 statistics can be confusing and that includes basic information such as the “infection rate”. This infection rate should be simply the number of people who have become ill with the disease and/or have tested positive for the antibodies, divided by the total population. Unfortunately, several things make this a slippery calculation. First, not everyone has been tested to determine if they are currently ill with the disease, or if they have built up any antibodies to the disease, indicating exposure and infection. Second, initial studies have indicated that the presence of antibodies tends to fade in exposed persons, ranging from a few weeks to a few months, complicating the ability for exposed persons to be counted if tested. Third, multi-generational households are more predisposed to contract and spread the illness due to proximity at home, or by the nature of their work. Lastly, the federal government is paying hospitals more for treating Medicare “Covid patients” than “non-Covid patients”, creating a conflict of interest for revenue-strapped hospitals. We are not trying to make a political argument, but we have read materials from scientists puzzled by the steep drop in mortality from ailments such as cancer or heart disease over the past 6-7 months.

Short Term Investment Performance. Beware of the “hot dot”. Many investment organizations will loudly trumpet the recent outperfor- mance of their hottest strategies, merely citing the last quarter or year’s attractive figures, showing their fund’s performance dot at the top of the peer group in advertising. While only one factor in the deci- sion-making process, we find it much more useful to focus on long-term performance (e.g. five- or ten-year average), identifying an investment strategy that consistently generates performance superior to its peers. Some funds are consistent “hot/cold” strategies, doing extremely well in some short-term periods and then poorly in the next. Each quarter, being the top performer in a particular asset class (or “hot dot”) is common, while being the consistent outperform- er is not.

Summary. Just like the oft-ignored parental advice to “turn in all your homework, because zeroes kill averages”, keeping in mind classic statistical traps can greatly deepen your understanding of a particu- lar topic. Don’t let someone bolster a weak argument, or chip away at your supporting data, as long as the numbers truly are well support- ed. As an organization, we continue to believe in the long-term viabil- ity of the U.S. economy, and promise to look deeper “into the num- bers”, so as not to be persuaded by a weak argument supported by poorly used statistics.