The Overreaction Hypothesis – Anatomy & How To Exploit It by Tom Macpherson, Dorfman Value Investments
In 1863 during the battle of Gettysburg, Union general Daniel Sickles decided to move his troops far in advance of the line demarcated by his superiors. As the troops moved out with flags flying and bands playing, another Union officer – Winfield Scott Hancock – watched the movement with great consternation. One of his staff officers commented that Sickles’ was disobeying orders. General Hancock looked through his binoculars, sighed, and said “they’ll come tumbling back soon enough.” And indeed they did. Hit with an enormous wave of Confederate soldiers, Sickles and his men suffered enormous losses (Sickles even losing a leg) almost causing the Union to lose the battle – if not the war.
SolarWinds (NYSE:SWI), had been roughly flat last year until – in the words of Hancock – it came tumbling back, losing 24.5% between July 15th and 17th, 2015. A tumble indeed. Looking at the quarterly report and revised outlook, I think the markets clearly overreacted to the company’s earnings call. I will get into the numbers in greater detail later in this article, but it would be helpful first to discuss why the market’s overreaction is important.
The Overreaction Hypothesis
A seminal work on this topic was published by Werner De Bondt and Richard Thaler in The Journal of Finance, July 1985. In their study, Bondt and Thaler placed the top performing 35 stocks over the past three (3) years into a portfolio entitled “Winners”. Correspondingly, the worst performing 35 stocks were placed in a “Losers” portfolio. They then tracked each portfolio’s performance against a representative market index for three years.
So what did they find? Interestingly, the “Losers” portfolio beat the markets handily while the “Winners” portfolio underperformed significantly. The cumulative difference between the two portfolios was roughly 25% over the three-year period. A case of inversion if I’ve ever seen one.
The essence of their findings was that for every overreaction in a stock price, there is an equal counter reaction. This works in both highly valued and low valued stocks alike. They described it as such:
“If stock prices systematically overshoot, then their reversal should be predictable from past return data alone, with no use of any accounting data such as earnings. Specifically, two hypotheses are suggested:
- Extreme movements in stock prices will be followed by subsequent price movements in the opposite direction.
- The more extreme the initial price movement, the greater will be the subsequent adjustment.”
DeBondt and Thaler suggested that the Overreaction Hypothesis represents the confluence of three effects:
- Directional Effect: Extreme movements in equity prices will be followed by movements in opposite direction.
- Magnitude Effect: The more extreme the initial price change, the more extreme the offsetting reaction.
- Intensity Effect: The shorter the duration of the initial price change, the more extreme the subsequent response.
SolarWinds: An Example
A case of the Overreaction Hypothesis is the stock SolarWinds (NYSE:SWI). This is a company with extremely strong financials. Return on Equity of 21%, Return on Capital of 23%, no debt, $542M on the balance sheet, and conversion of 47% of revenue into free cash. Management has grown revenue at 21% annually since 2005 and free cash at 14% since 2010.
On July 16th, 2015, management reported Q2 numbers. In addition to missing Q2 revenue estimates (while beating on EPS), the company guided for Q3 revenue of $130M-$134M (below a $136.1M consensus) and 2015 revenue of $502M-$512M (below a $519.7M consensus). EPS guidance was better: $0.49-$0.53 for Q3 (consensus is at $0.52) and $2.00-$2.08 for 2015 (consensus is at $2.00).
Anatomy of an Overreaction
These results provoked a collapse in the stock price. The shares dropped to $35.54 on July 17th, 2015, down 24.5% from their close of $47.05 on July 16th, 2015. I was surprised by the reaction of the markets. Solar Winds reduced Q3 revenue guidance by 4.5% but was still projecting 15%-19% Y/Y growth. 2015 revenue estimates were reduced by 3.5% but still projected 17%-19% Y/Y growth. Using a discounted cash flow (DCF) model, these changes reduced our estimated fair value from $54/share to $53/share or an estimated decrease of less than 2%. Yet that big one-day loss suggested that SWI was worth roughly 25% less than it had been a few days earlier. Followers of Efficient Market Theory (EMT) would tell me the price decrease represents the best information available and the true value of the company’s shares. I think this type of thinking is silly and a gross overreaction to the earnings call.
Following the massive downside movement, according to Bondt and Thaler we should have expected to see a short-term, significant upturn in price. In my July 21, 2015 article on Guru Focus, I wrote, “That may be. But even if this doesn’t work out, I have in my back pocket the ultimate asset for the value investor – an outstanding company selling at a significant discount to our estimated intrinsic value.” I added, “Whether SolarWinds is a successful example of the Overreaction Hypothesis will be tested over the next several months.
So how did their model work out? After dropping from $47 per share to roughly $36 per share on July 17th, 2015, the stock slowly rose to $41.50 through Oct. 8th, 2015. On Oct. 9th 2015 the stock jumped nearly 14% after the board announced it was exploring strategic options including the possibility of an LBO or acquisition. Talk on the Street mentioned discussions with PE firms and a sale price of roughly $55 to $60 per share. Then, on October 21, 2015, the company announced it was being acquired by Silver Lake Partners and Thoma Bravo for $60.00 a share. In this instance, Thaler/De Bondt’s hypothesis of a rapid, substantial and upward price move emphatically did take place.
How to Exploit the Overreaction Hypothesis
So do the data guarantee an equal and proportional response for every sharp price movement? No. But Thaler/De Bondt’s research shows it happens enough that I can make some general observations going forward.
First, I can assume a considerable price movement gets enough animal spirits moving that the markets can see similar reactions to Thaler/Bondt’s research. Second, these animal spirits are often driven by emotions rather than sound qualitative thinking. This means the basis for these market reactions is less about data and more about passion. Last, profound price changes can provide incentives for varying types of investors. A significant drop may produce interest in activist hedge fund players. So it shouldn’t be that surprising when market reactions frequently counteract a significant pricing event.
That said, I would posit there are several conditions that might enhance an investor’s chances of achieving positive results from such dramatic price moves.
Price movements push valuation extremes: In the case of SolarWinds, the dramatic drop in price pushed the stock’s valuation into highly undervalued territory. I think investors would be rewarded most when the price action pushes valuations into highly over or undervalued territory. These valuations would greatly enhance the chance of Mr. Market making a strong counter move.
Quality can impact your outlook/position: I believe investors would be wise to look for sudden downward price moves with high quality companies (in this case go long) or the opposite for low-quality stocks (short the stock after sudden upward price moves). In the case of SolarWinds, I believe the market saw a high-quality company trading at a very low price – hence the surge upwards.
Always, always focus on valuation: Ultimately investors have no idea about the catalyst in the Overreaction Hypothesis. As value investors we are best to focus on the impact of the sudden price move on valuation, and ultimately let the chips fall where they may.
Thaler and De Bondt’s research has stood the test of time since they first published it roughly 30 years ago. Data tell us we can expect – in general – a reaction similar to our experience with SolarWinds. Whenever I see a case of the Overreaction Hypothesis I take a long hard look at the company, the reasons behind the pricing event and the relationship between the new price and estimated fair value. While I never assume I will see a reaction as outlined by Thaler and De Bondt, I’m certainly aware it could happen – and waiting for it with open arms.
This article is adapted from columns on Gurufocus.com July 21, 2015 and October 13, 2015