The Case of Civeo and the Unknown Unknowns

A wildfire in Fort McMurray, Canada, destroyed over 2,400 homes in the first two days of a furious burn that spread across the region and forced the evacuation of thousands.

But the fire’s impacts were more positive for one leveraged small value stock called Civeo.  Civeo provides remote site accommodations to the North American resource industry.  As of March, Civeo's vacancy rates were soaring and prospects looked bleak.

The fires changed that. The company struck a deal with the government to provide housing for the wildfire’s refugees.  Civeo's share price spiked about 30% in reaction to this surprising turn of events (the graph below shows the stock price relative to Google searches for “Fort McMurray fire.”)

Figure 1: Searches for Fort McMurray Fire vs. Civeo Share Price

Source: Google Trends, CapitalIQ

What makes this story so interesting is its unpredictability.  No financial analysts included wildfires in their models. No investors drew event trees and assessed probabilities of wildfires as they developed their investment models for Civeo.  This was an unknown unknown.

And this particular unknown unknown makes blazingly clear a central truth of investing: surprises happen far more frequently than we would expect, and these surprises are the prime driver of share prices. (This is true in life more broadly: how many of the most important events in your life were predictable a year or two in advance?). 

Acknowledging the importance of surprises does not mean that active investing is as fruitless an exercise as hiring a fortune teller, however.  Acknowledging unpredictability simply changes the goal of the investor, from making superior predictions to choosing the stocks that are most likely to benefit when surprises unfold. Investing is a game of meta-analysis, not analysis, and to beat the market we need only identify stocks most likely to exceed the market’s expectations.

Choosing stocks most likely to benefit from future surprises is, fortunately, much easier than making better predictions about the future.  In 1997, a dream team of financial academics co-authored a paper that proved empirically that some stocks were more likely to benefit from unforeseen events.

Rafael La Porta, Josef Lakonishok, Andrei Schleifer and Robert Vishny titled their paper “Good News for Value Stocks,” a clever play on the idea that news would more often be good than bad for certain types of companies.  They found that a significant portion of the return difference between value stocks and high-priced glamour stocks is “attributable to earnings surprises that are systematically more positive for value stocks.” Below is the table showing their key findings:

Figure 2: Key Findings from "Good News for Value Stocks"

Source: "Good News for Value Stocks"

Surprising news disproportionately benefits the cheapest stocks in the market (those the market is most pessimistic about) and punishes the most expensive stocks (those the market is most optimistic about).  Civeo has experienced the impact of surprises in both directions.

Civeo was once a high-flying growth stock with a great story.  At peak market capitalization in early 2014, the company’s public ownership read like a who’s who of prestige managers: JANA Partners, Capital Research and Management, Greenlight, Royce, Brigade, and many others owned large stakes.
Sterne Agee initiated coverage in June of 2014 with a price target of $24, predicting that Civeo would earn $375M of EBITDA in 2015 and trade at an 11-12x multiple.

We all know what happened next.  Oil prices crashed and Civeo’s stock price fell from over $25 to less than $1.  And the prestige investors who bought with a thesis like Sterne Agee’s sold their shares at some point during the long drop, long before the vulture investors like Verdad bought in.

Figure 3: Oil Prices & Civeo Share Price

Source: FRED, CapitalIQ

Fast forward to late 2015 when Verdad’s linear regression model identified the stock as an attractive opportunity.  The company made $123M of EBITDA in 2015, far from Sterne Agee’s forecast.  We bought in $2.12, which valued the company at 3.4x EBITDA.  At this valuation, the company was earning an 87.8% LTM free cash flow yield and 66% leveraged.  I drew up the following probability tree for my estimate of what Civeo was worth at different oil prices, and the base rates of those historical prices.

Figure 4: Probability Tree for Civeo 

Dirt cheap, highly leveraged, cash generative, paying down debt. The company had everything from a quantitative perspective.  Applying a reasonable probability weights to potential different outcomes, the stock seemed like a decent bet within a broader portfolio.

But oil prices were well below the break-even for North American production, vacancies at Civeo locations were spiking, and there seemed no prospect for immediate recovery.  At current oil prices, the stock was most likely worthless.  

That was until the fires broke out.  And an unforeseeable event drove a major recovery in the share price.  Thomas Wolfe famously observed that the course of history hinges on the most surprising and trivial occurrences.  “Each moment is the fruit of forty thousand years,” he wrote. “The seed of our destruction will blossom in the desert, the alexin of our cure grows by a mountain rock, and our lives are haunted by a Georgia slattern, because a London cutpurse went unhung.”

In Civeo’s case, the alexin of the cure grew in Centennial Trailer Park, and the stock’s price was haunted by an Athabasca oil sand explosion, because rain hadn't come.