How Richard Burr Made $185,000 During the Covid-19 Pandemic
Senator Richard Burr (R-NC) is the focus of an inquiry into major stock sales in advance of the February-March stock market crash by senators who received closed-door briefings on the spread of Covid-19. As head of the Senate Intelligence Committee, Burr received daily confidential reports throughout January and early February on the potential impacts of the coronavirus pandemic. ProPublica reported that on Feb. 13 – less than a week before the market peaked on Feb. 19 – Sen. Burr sold between $628K and $1.72M of his stock in major companies. According to our estimates, this sale amounted to about 25% of his financial holdings at the time. The FBI is currently investigating Burr for violating the 2012 STOCK Act, which prohibits members of Congress and other government employees from trading on non-public information. Sen. Burr has temporarily stepped down as chair of the Senate intelligence committee in response to the scandal, but maintains that his trades were exclusively informed by public media reports about the pandemic.*
It’s hard to know what information Burr used in making his trades. One approach — taken by most journalistic outlets — involves piecing together a timeline of Burr’s information flow and communications immediately before executing the suspicious trades. Another approach, sometimes taken by the Securities and Exchange Commission (SEC) when prosecuting insider trading on Wall Street, relies on evidence that traders with inside information enjoyed massive and unusual financial returns from their transactions. Evidence that Burr’s stock trades resulted in extraordinary returns would make his claim that he was trading based on purely public information less convincing.
Data aggregated by GovTrades gives unique insight into this second approach by directly quantifying the financial payoff that Burr enjoyed from his February stock sales and studying how unusual the returns actually were. To do so, we pulled Burr’s financial holdings and stock market transactions from January 1, 2014 to May 22, 2020, along with associated daily market returns.
We find that Burr earned a significant sum from his February sales — over $180,000 over less than 100 days, equivalent to a 25% return. This amount is about three-and-a-half times the annual income of the median North Carolina family he represents. Burr’s previous stock trades rarely earned anything close to that amount. To achieve such high returns, Burr selectively sold shares in companies most impacted by government-mandated closures and stay-at-home orders – even though such policies had yet to be announced to the public. Finally, neither Burr’s defense for his trading behavior, nor alternative innocent explanations, cannot explain his astronomical returns. While not definitive, our analysis adds crucial evidence in support of those who claim that Burr abused his office by trading on non-public information.
1. Burr’s February trades earned exceptionally high returns compared with past trades.
How much money did Burr earn from his February trades? To find out, we estimated the value of Burr’s actual financial holdings since Jan 1, 2020, and compared the result to an alternative scenario in which Burr did not sell any stock on February 13. Figure 1 plots the daily value of Burr’s actual holdings (red line) and the value of his holdings had he not sold his stock (blue line). As of May 22, 2020, Burr saved almost $187,000 by selling at the exact right time. That’s equivalent to a 25% return in less than 100 days, or over 65% in annualized terms.**
This striking return should be enough to raise eyebrows. It also stands out in the context of returns from Burr’s previous financial transactions. Fig 2 shows the full distribution of 100-day returns*** for each of the 121 trades recorded in our database from Jan 1, 2014 to Feb 13, 2020. Over the past 6 years, the median return was a whopping 0%. The red line on the graph plots the return from the most profitable Feb 13 sale (in Park Hotels & Resorts, Inc, NYSE:PK), which exceeded the return from every previous stock transaction. The green line indicates the average return for all Feb 13 sales – higher than all but 8 previous individual stock transactions. Burr’s uncharacteristically high returns suggests he may have had an informational edge in his February trades.
2. Burr sold stocks especially impacted by the Covid-19 pandemic.
Sen. Burr avoided such stark losses by selectively selling companies in his portfolio that were disproportionately impacted by the pandemic. Fig 3 lists ten stocks from Burr’s sell-off with the largest loss in market value since Feb 13.
These sales indicate that Burr anticipated a slowdown in tourism and travel; visits to bars and restaurants; and general business activity. Specifically, his Feb 13 transactions include:
A $150,000 sale of stock in two hotel chains – Park Hotels & Resorts and Wyndham Hotels and Resorts, both of which experienced a big drop in number of visitors
A $50,000 sale of Constellation Brands stock, an alcohol supply company with profits driven by sales to upscale bars that may struggle to offer takeout and delivery
A $150,000 sale of stock in two companies that produce inputs for vehicle manufacturing – Orion Engineered Carbons, which produces a major component in tire manufacturing, and Axalta Coating Systems, which manufactures coatings for motor vehicles. Vehicle sales have fallen substantially since the start of the pandemic
An $8,000 sale of FedEx stock, which has experienced a big decline in shipping demand due to lagging economic activity
By selling some of his worst-performing stocks, Burr’s Feb 13 transactions improved the overall return of his financial portfolio. Fig 4 compares our estimate of Burr’s actual portfolio returns**** to counterfactual estimates from the previously-referenced alternative scenario which freezes holdings at Jan 1, 2020 values, excluding Feb 13 sales. By May 22, the actual portfolio return for financial assets was about 2% higher than it would have been had Burr held onto all of his assets. This suggests that Burr’s Feb 13 sales did not represent random, panicked selling in response to unspecified anxiety regarding Covid-19’s impact on the economy. Rather, the transactions seem calibrated towards sectors negatively impacted by the imminent government stay-at-home orders, restaurant closures, and travel advisories that had yet to be announced to the public at the time.
3. Burr’s excuses can’t account for his high returns – and neither can other innocuous explanations.
In his defense, Burr claims that he made his trades based on publicly available information. However, Burr’s trades were completely out of line with public market sentiment. Fig 5 plots Burr’s daily transaction volume since Feb 1, 2020 as a fraction of the average value of his stock transactions since 2014. For comparison, Fig 5 also plots average daily S&P 500 transaction volume as a fraction of average transaction volume since 2014. Bars above the horizontal line at one indicate above-average daily transaction volumes, and bars below the line indicate below-average daily transaction volumes. Sen. Burr’s stock sales – over twelve-times larger than his typical stock transaction – came at a time of relative market calm. Had Burr been trading on public information, it is likely that others would have sold their stocks around the same time.
Clearly, Burr had a much different interpretation of public information than other market participants. However, in this context, there are several potentially innocent explanations for his trading behavior and unusually high returns. Unfortunately, the data do not support any of them.
First, Burr may just be smarter than the rest of the market. Fig 6 dispels this notion by comparing Sen. Burr’s portfolio returns since Jan 1, 2014 against the S&P 500. Burr’s portfolio significantly underperformed the S&P, earning about 40 percentage points less over this 6-year period.
Second, Burr may have been particularly attuned to the potential economic consequences of a pandemic, and has always structured his financial portfolio accordingly. Indeed, Burr authored the 2006 Pandemic and All-Hazards Preparedness Act and has consistently emphasized vigilance against pandemic threats as a senator. In that case, we would expect Burr’s portfolio to perform relatively well when put to the test during a real pandemic. However, as Fig 7 shows, Burr’s portfolio returns also underperformed the S&P 500 since the onset of the Covid-19 pandemic. So it seems like Sen. Burr’s financial decisions have not historically been informed by pandemic preparedness – at least, not until after he began receiving confidential briefings about Covid-19 in the US.
Third, Burr might be especially good at trading right before periods of high market volatility, partly explaining the extremely high returns from his Feb 13 trades. To investigate, we calculated the standard deviation of returns in the 90 days following every trade that Sen. Burr made since Jan 1, 2014. Fig 8 plots this basic measure of market volatility against the 100-day return from each previous stock transaction, along with a trend line. The red dot plots the average volatility for Feb 13 trades against their average return. There’s basically no relationship between leading market volatility and financial returns, indicating that the pandemic-induced market turmoil did not present some opportunity that played to Sen. Burr’s unique investing skill set.
4. Has Burr engaged in suspicious trading before?
The timing, targets, and returns of Burr’s February sales all suggest that Burr may have had an informational advantage. But there have been previous instances where Burr has made similarly striking short-term returns from his stock trades. The fact that these trades earned higher returns than the suspicious Feb 13 sales suggests that they may have also been partly based on inside knowledge. Figure 9 zeros in on the 8 trades since 2014 that had a 100-day return at or above the average for Burr’s Feb 13 sales.
If we’re looking for other suspicious trades by Richard Burr, these would be an excellent place to start.
*****
Of course, we cannot completely rule out the possibility that Burr did in fact transform from a mediocre retail investor to a stock market guru overnight, or that he just got lucky. However, we believe that our market-based approach to analyzing Burr’s February sales offers a useful complement to the more narrative investigations undertaken by other outlets. Taken together, alternative explanations for the data may look more and more like wishful thinking.
*****
* The STOCK Act does not prohibit trading based on publicly available information.
** To estimate percentage returns, we take the gross value of Feb 13 transactions, subtract the counterfactual market value on May 22 of these holdings, had they not been sold on Feb 13, and divide by the gross value. This calculation assumes that sales were held in cash at 0% return from Feb 13-May 22. We then compute the compound annual growth rate to annualize this return.
*** For purchases, transaction returns are calculated using the percentage growth in stock price from the purchase date to the nearest business day 100 days after the purchase date. This assumes that funds used to make the purchase would have been held in cash had there been no transaction. For sales, transaction returns equal the negative of this percentage growth. This assumes that sales are held in cash after the transaction.
**** “Portfolio returns” calculated in this section differ slightly from the returns referenced earlier. Portfolio returns indicate the daily returns of assets currently invested in the market, while returns calculated in other sections compare purchases / sales to holding cash.
