Self reflection is not always comfortable, but it can be beneficial for mental well-being or when it comes to learning from and improving upon past decisions and ultimately becoming a better human—or at least striving to be.

Fund managers often disregard self reflection, for a number of reasons. Sometimes, it’s pride. Sometimes, they don’t have the tools necessary to help them reflect on investment decisions they’ve made. Other times, it’s the institution itself that hinders this type of personal development. Take, for example, a long-only fundamental fund. Since these types of firms typically don’t often trade, is it beneficial to analyze their data? 

Clare Flynn Levy, founder and CEO at Essentia Analytics, says the assumption is that data analytics won’t work for that type of fund. “An objection we often hear when we’re talking to potential clients is, ‘I have a few positions in my portfolio and I don’t trade them very much. Doing data analytics on that is sort of an interesting one-off exercise, and then the data is not going to change that much because I don’t trade that much.’ That’s the assumption.” 

She explains that there’s still value in analyzing data about decisions a fund manager makes that don’t result in trades. 

“Your job is to be a decision-maker, but no one’s capturing the decisions you’re making not to trade, so how do you know that those are good decisions or not? You don’t,” Flynn Levy says. “Somebody who’s trading a low-turnover, concentrated portfolio needs to be able to reflect, as much as somebody who’s trading high-turnover in a gigantic number of positions portfolio.”  

One of the funds Essentia worked with—”Fund A”—is a $1 billion equity portfolio within a $50-plus billion investment management firm, managed by two portfolio managers. Fund A is a highly concentrated fund holding 30 positions and has an annual turnover of less than 5%. 

Fund A implemented three of Essentia’s tools—Brain Dump Nudge, Alpha Decay Nudge, and Vulnerable Positions List Nudge—which led to an overall improvement of an additional 600 basis points, or $60 million of additional profit on a $1 billion fund. 

Essentia set up a daily Brain Dump Nudge for Fund A to capture the manager’s thoughts as to why they traded—or why they didn’t. Flynn Levy explains that this nudge would prompt the managers daily with three questions: Which stock do you wish you owned more of, why, and why are you thinking about this stock? The third question has five answers correlated to non-earnings news flow, earnings reports, research, and due diligence, and managers can select all that apply. 

“Once a day, it’s capturing some data about what they’re thinking, and getting into that habit and building a dataset that can be analyzed down the road,” she says. “To the extent that they do trade, we can then take trade data and combine that with the data about what they were thinking about, and connect the dots between.”

They implemented a monthly Alpha Decay Nudge to remind the managers to check on positions they may be at risk of holding on to for too long based on previous patterns. Essentia also implemented a weekly Vulnerable Position Nudge to help decide whether to exist losers sooner. 

Fund A began using Essentia in November 2016, by first doing a bulk import of their historical trade and holdings data, and then with an ongoing daily feed. 

Usually, Essentia looks for at least 2,000 historical trades as a starting point to find statistically significant insights. Fund A didn’t have a long trade history, but it could still be analyzed. 

“They didn’t have [2,000 historical trades] yet because their fund was relatively new, but we can still analyze what they have. Now we’re just not going to necessarily say anything conclusive about the pattern, but you could look at yourself in a distant mirror and you can still recognize yourself and you still see, ‘Oh, I have stuff on my shirt’ or whatever—you can still see it. What resonated with them was they could see that they were making all of their alpha performance out of their biggest positions, and they were consistently destroying value in their smaller positions,” she says.

Too Long-Term?

The problem is the way that some portfolio managers are trained at the very start of their careers. Before founding Essentia, Flynn Levy spent 10 years as a fund manager. She, too, felt the pressure of being long-term. 

“Every fund manager is taught that you’re supposed to be long-term. If you’re a trader, you’re not—you’re trying to make short-term gains—but a fund manager, typically, it’s all about being long-term,” she says.

However, there is a risk of being too long-term. Flynn Levy explains that markets change and companies change, and sometimes it’s pure chaos—as in the case of the Covid-19 pandemic. 

“If we’ve learned anything in 2020, it’s that you don’t know what’s going to happen. You can’t just necessarily assume that every company you own is going to last for the next 20 years,” she says. 

The important thing to figure out is where they tend to stop adding value in the lifecycle of positions. “There are four main profiles that emerge and the most common one is the round trip where you generate lots of alpha in a position in the first half of its life, and then you give it all back, which is a good explanation for why active fund managers have not been performing net of fees,” she says. 

For example, it could be that the average position in a portfolio tends to run out of “alpha juice” by month 33. “I’m not saying that every time a stock in your portfolio gets to 33 months old, you have to sell it, because that past isn’t necessarily predictive. But, month 33 would be a good point for us to nudge you and ask some questions about [your holdings],” she says. 

In Fund A’s case, Alpha Decay Nudge was set at months six and 24. These periods represent the managers’ particularly effective decision points to review a stock’s investment thesis and question if it still offers the opportunity for alpha generation. 

Essentia uses the Nudge’s structured data to feed its model, meaning that the “why” doesn’t go into its model. “Of course, the technology is increasingly there to analyze text, but the way a fund manager writes, it’s not the same for every person; it could be shorthand,” she says. 

That said, the “why” portion is still valuable for two reasons. “Dumping your brain” in that way is like journaling, Flynn Levy explains. “It’s useful for freeing up your cognitive bandwidth. So, managers can just write it down for posterity. … It’s healthy, but also what happens is that you end up being able to see, later on, what you were thinking,” she says. 

On the flip side, Essentia has also dealt with high-turnover funds. One recent example is “Fund B”—a $150 million long/short hedge fund within a multibillion-dollar family office. Fund B had over 3,400 trading decisions over 18 months. These were imported into Essentia’s analytics engine, and a secure feed was established to upload new data daily. The manager for Fund B suspected that trimming positions into earnings reports—one of the manager’s heuristic techniques—no longer generated alpha for the fund. 

The data showed that Fund B’s manager was a strong stock picker both on the long and short sides. Flynn Levy says this proves that not all of Essentia’s insights are negative. However, the insights showed that the fund manager’s trimming behavior was destroying more than 700 basis points of value per year. A third of that was particularly when the manager was trading ahead of earnings reports, as he suspected. 

Essentia implemented a daily Brain Dump Nudge asking the manager what the upcoming catalysts are in the portfolio and whether he had done sufficient research to make decisions ahead of those catalysts. Essentia also created a new Add/Trim Context Nudge, prompting the manager with questions like whether the research process is complete whenever he trades a position he already holds. 

As a result, the manager’s trimming behavior went from negating 723 basis points annually to 229 basis points—an improvement of almost 500 basis points. Overall, his performance improved by 553 basis points per year, or $8.3 million profit on a $150 million fund. 

“The moral of the story is, for portfolio managers who are brave enough to look in the digital mirror, there are massive performance gains to be made—we call that ‘behavioral alpha,’” Flynn Levy says. 

Looking ahead, Essentia is developing a new capability to allow portfolio managers to break down their analysts’ performance statistics by sector. For example, for a long-only equity fund, Essentia analyzes it against the MSCI World Index. 

But the fund manager might also ask for a breakdown of that analysis by sector because their analysts are assigned to specific sectors. Showing that by sector isn’t an issue, but it could make for an unfair comparison of analysts’ performance. 

“If the energy sector has been a massive underperformer in the MSCI World for years, which has been the case, the energy traders are going to look really bad because they can only pick energy stocks. So that’s not really a fair comparison to make about whether they’re doing their job well. What you want to do is compare them to the MSCI World Energy Index,” she says.

The new capability is currently being coded and will be available to users in the first quarter of 2021.