March was a bloodbath for most investors. The market fell heavily and hidden beta exposure revealed itself within investment strategies. Meanwhile, traditional quant factors collapsed with unusually high correlations only seen during panics.
Yet those with access to 2iQ global insider transaction data could have outperformed. Here at 2iQ Research, we monitor insider buying and selling closely, and throughout March insiders transacted in record numbers. Indeed, the number of transactions was often twice what we usually observe, as highlighted in the Bloomberg article Corporate Insiders Send Most Bullish Stock Signal Since '99. What’s most interesting about this increased level of insider transactions, however, is that insiders demonstrated their level of insight, delivering record long-short alpha performance over the month.
We have known for a long time that insiders have deep insight and are some of the most informed participants in the market. That’s why, in 2015, we developed a model to track the best names as a function of the underlying insider transaction data. We have tracked this model’s performance for more than four years now, reporting the monthly performance to our subscribers. Given that March was an exceptional month, we will take a closer look at the data.
A record month
Our EnhancedRatio3M model, which has been live since September 2015 delivered record returns both across regions and globally for the month of March.
We have used a simple methodology to track the performance of the model since it went live in September 2015. We have taken the top 20% long names and the bottom 20% short names and then rebalanced monthly on the S&P BMI Universe. Returns have been risk adjusted using a cross-sectional risk model.
March 2020 was the strongest performance at the global level in and out of sample, since we launched the model. It was the best performance for North America, and the second best for Europe and the Asia Pacific. The three regions outperforming together is a testament to the orthogonality of the model, and highlights the power of insiders’ insights, particularly in turbulent markets.
At the global level this was definitely the best month for insider transactions and our model. We were happy with the previous 4.5 years of out-of-sample performance, as it provided value consistently across regions. Yet being able to provide outstanding alpha in an extremely turbulent market is a fantastic achievement.
What drove the performance of the model?
During March our model was very effective at differentiating names within the Healthcare, Financials, and Energy sectors (sectors with the best selection effect).
Here are a few examples of companies that were positioned well due to insider transactions and our model:
- Celltrion Healthcare Co Ltd (068270.KS) - KOR - Healthcare
- Logitech International SA (LOGI.US LOGN.SW) CH - Information Technology
- Conagra Brands Inc (CAG.US) USA - Consumer Staples
- Astrazeneca Plc (AZN.LN) GB - Healthcare
- Amazon.com Inc (AMZN.US) USA - Consumer Discretionary
And in the small-cap space we had a long indicator on:
- Inovio Pharmaceuticals Inc (INO.US) USA - Healthcare: +76%
- Apache Corp (APA.US) USA - Energy
- Occidental Petroleum Corp (OXY.US) USA - Energy
- Petróleo Brasileiro SA (PBR.US) BR - Energy
- Scentre Group (SCG.AU) AUS - Real Estate
- Amarin Corp (AMRN.US) IRL - Healthcare
Ultimately, the model was able to capture alpha in the large-cap space as well as in the small-and mid-cap space, across a variety of sectors and regions.
Insiders are not like the other factors
Two ‘factors’ that performed poorly in March were Value and Short Term Momentum (Reversal). Given that our insider transaction model has a relatively low correlation to these traditional factors, this helps to explain why it outperformed in March.
As MSCI analysts said in a recent blog post: "Among alternative data sources we examined, the insider-transaction dataset stood out with unique characteristics when compared to traditional risk-model style factors."
Chart 3 highlights the low correlation and the orthogonality of insiders’ behaviour especially after being filtered by our model, which adjusts transactions for the usual insiders’ biases. Insiders do have a bias towards value so our model generally reduces this exposure by 5%-15% over simpler metrics like CountRatio3M highlighted in the chart above. Our methodology therefore limits the correlation between our insider model and the Value factor.
March was a smooth ride up
In Chart 4, we track a daily rebalanced portfolio using our EnhancedRatio3M signal. Performance varies slightly from the monthly holding period but gives a good indication of the level of stress and the drawdown that the strategy experienced during the month.
Chart 4. Daily Rebalancing, EnhancedRatio3M L/S Performance
Both within regions and globally the trend was mostly up and relatively smooth. We noted that Europe and Asia Pacific peaked around the middle of March. All strategies rallied in the last week of March as the market stabilised a little and investors were able to differentiate between the winners and the losers of the previous few weeks of turmoil. Clearly, insiders demonstrated their knowledge during this period.
As you can see, the information provided by the market’s most informed participants can be extremely valuable. It is uncorrelated to other more traditional factors, even during periods of market turmoil.
There is no doubt that further volatility is awaiting us. We could see further downside before a sustained recovery takes place. We believe that in times like this, insider transaction data for cross-sectional stock selection is an indispensable tool for investors whether you have a fundamental, quantamental, or fully-systematic approach.