W.K. Kellogg Foundation Using Machine Learning To Generate Real Alpha | Joel Wittenberg, VP & CIO | Q&A
Joel Wittenberg is the vice president and chief investment officer at the W.K. Kellogg Foundation, where he has served since 2009. In this interview, he tells us about the Foundation's project that focuses on funding emerging minority-led firms; why they're implementing machine learning technology in their investment office; and how the investment team’s 9-year history and diversity are key drivers to their success.
Joel was recently named as one of Trusted Insight's 2018 Top 30 Foundation Chief Investment Officers. This interview has been edited and condensed.
Trusted Insight: It’s been some time since your last interview with us. What are some of the more notable changes at the Kellogg Foundation?
Joel Wittenberg: It’s been an exciting three years at the Kellogg Foundation. The Foundation has moved to more of a team-based approach, also called a networked organization model, to improve the delivery of our mission goals of thriving children, working families and equitable communities. While children remain at the center of all we do, they need to live in stable families with quality jobs, and in equitable communities. Working across these interconnections is how we seek to help all children thrive.
"Our goal is to identify firms starting out, people that maybe had a track record somewhere else and left to start their own firm. These firms have diversity through African American, Latino, Asian, Native American or female ownership."
On the investment side, we’ve also made several changes as well over the last few years. We've been much more focused on our hedge fund portfolio driving a combination of higher volatility across managers with lower correlations. On the private equity side, our portfolio has performed significantly above benchmark so we’ve used this opportunity to raise the bar on our managers as well. We've also upgraded our public equity managers. As you can see, we’ve used our strong momentum to drive us ahead.
Trusted Insight: In your last interview with us, you mentioned a project focused on funding emerging minority-led firms. How has that portfolio developed over the last few years?
Joel Wittenberg: That's been a great and exciting process for us. As I said, the Kellogg Foundation focuses on racial equity as a key part of this mission. That puts us in a great spot to develop relationships with newer, diverse-owned firms. Our goal is to identify firms starting out, people that maybe had a track record somewhere else and left to start their own firm. These firms have diversity through African American, Latino, Asian, Native American or female ownership. We currently have about $120 million in the program and have identified several managers to receive much larger allocations or expected future allocations, which we view as a huge success. On a performance basis, we’ve been very happy with the investment returns of the program and I expect this program to continue to identify managers for the future who can eventually take on larger mandates.
The interesting thing about this program is that these firms are brand new. They're just starting out so many of them have difficult times. We’ve had a couple of these firms that haven't made it. The key is having our board’s support for this strategy because, in the past, I'd have been pretty upset if I hired a firm that then didn't make it. It allows us to take more firm risk in this portfolio, because, as I said, we have the support of our foundation board.
Trusted Insight: About 7 percent of the portfolio is in venture capital. In the last decade, the price of starting companies decreased significantly to $100,000 or less. What’s your strategy for cutting through the noise in this highly competitive and crowded market to find managers that can provide healthy returns?
Joel Wittenberg: We're seeing VC funds we invested in a decade ago, that had to make multi-million dollar investments into startups which took years to get going. Today, it can often take $100,000 and six months to get it going. That gives any newer fund a significant competitive advantage.
"We're using machine learning to generate key drivers with our data, and artificial intelligence provides us the opportunity to improve on our performance."
We're very focused on some of those ideas that have a very short runway, without millions of dollars and years to implement. About 4 years ago, we started with some funds that used this model and we're seeing great returns from those funds. In addition, we're invested in funds through TI that are doing this successfully. It's been a real competitive advantage.
Trusted Insight: Is the Kellogg Foundation using in-house technology to help with the investment process? If not, is it something that you’d consider given the rapid development of machine learning?
Joel Wittenberg: We're actually thinking about how to apply a machine learning model to our asset allocation that would allow us to make some tilts in the portfolio. We haven't implemented it yet, but it's funny you say that because I was just meeting with my team on an update. We believe we have an opportunity to add some real alpha that can be generated from the movement between liquid allocations based on what the model tells us. We're using machine learning to generate key drivers with our data, and artificial intelligence provides us the opportunity to improve on our performance.
Trusted Insight: Do you think other technologies like blockchain will make its mark in the institutional investing industry?
Joel Wittenberg: We're excited about blockchain. We believe that it's going to create a lot more transparency in the marketplace. For example, we sold some of our private equity on the secondary market. There’s poor transparency in that market. When blockchain becomes the norm for secondary private equity transactions, our transparency will improve, which will improve the bid-offer spreads significantly.
We're also excited about the banks implementing blockchain for custody, which will give institutional investors the ability to really drive our custody costs lower. We’ll have better transparency into their cost structures. We think blockchain is going to be a real asset. I think it's going to be a big advantage for us in reducing our costs. Of course, we’ve also seen investment into blockchain within the portfolio as well.
Trusted Insight: Are there any areas of uncertainty that you’d like to focus on some more?
Joel Wittenberg: It's really back to the machine learning and AI that we're thinking about. We are looking at it from two additional perspectives. How do the investments that we make utilize it? And we're trying to identify long-only, hedge fund and PE managers who utilize machine learning and AI to identify appropriate equities. We're looking for investments in companies that are making the transition in their decision making support to include machine learning and AI.
It's amazing the number of sophisticated Fortune 500 companies that don’t yet understand or haven’t implemented AI and machine learning. They must first understand it and then embrace it to identify drivers and decisions if they are going to continue to succeed. We think that the ones who understand it and have implemented it now have a very big competitive advantage. We want our public equity managers digging into that information and identifying who's using it effectively.
Trusted Insight: In what ways is the Kellogg Foundation unique in comparison to its peers?
Joel Wittenberg: The investment team is the key driver of our uniqueness.
My investment team has been together now for about nine years. We have a total of four directors. The three directors that handle our diversified assets have been together for nine years. Our fourth director handles the mission-driven portfolio. The benefits of this stability can’t be replicated and allows us to stick to our well-designed plan.
By sticking to the plan, we’re not making constant mid-cycle changes, which of course always happens at the worst times. The team enjoys being together and we all have a great understanding of each other, where each other comes from and how we think about investing. We know each other’s strengths. We know each other’s weaknesses. That's a huge advantage that we have in identifying, debating and implementing ideas. We’ve created investment and approval processes that have proven to be real alpha generators. We are nimble and respond to opportunities quickly using our seasoned processes. Our stability is a competitive advantage that can’t be replicated. I’m so grateful to all my team for their hard work and contributions to the foundation investments and the foundation’s mission as well.
Second, we're all generalists here at the Kellogg Foundation, and each of the team members has unique capabilities. One person is focused on our hedge funds, although he's still a generalist. Another member is focused on our venture capital. And, another member is focused more on our public equities. But as generalists, we can look across asset classes to identify real value rather than filling up allocation buckets.
Third, our diversity is a competitive advantage. The team comes from all sorts of backgrounds. The investment team reflects the Kellogg Foundation’s fabric when it comes to diversity, and it's something that clearly adds value. Last year, we finished our best three-year period in the team’s 9-year history. I am convinced that that three-year performance is driven and comes from the diversity of thoughts and backgrounds that we have within the group, as well as our stability of 9 years together. As I said, it's just a huge competitive advantage and shows that diversity is also profitable.
Last, I would emphasize that with $3.5 billion of diversified assets and a strong governance process, the Kellogg Foundation is in a position that makes us large enough to be important to a manager, but not too large that we can't be nimble. We're finding interesting idiosyncratic ideas out there in smaller and newer funds that allow us to invest and maneuver into those funds where large peers can’t get the right allocation. I believe that our size puts us really in a sweet spot of how we go about allocating.
Trusted Insight: We recently spoke with a CIO at a prominent Foundation, and he spoke of the same concept. He was at a super-large institution and then transitioned to a smaller one, which brings more flexibility and nimbleness in generating attractive returns.
Joel Wittenberg: Yeah, that's exactly it. In addition, the foundations and university endowments of a size like the Kellogg Foundation share a lot of these idiosyncratic ideas together, which really makes it a lot of fun. We know that other smart people are doing due diligence on things that are similar or sharing ideas because we know each other. It’s funny, some of the top funds are located in the state of Michigan and we are all really good about sharing our ideas, successes, failures and lessons learned. We have a great and collegial group of allocators here. It's a nice network.
Trusted Insight: Is there anything that we failed to ask about you, Kellogg or foundation investing broadly?
Joel Wittenberg: We all joined the foundation for various reasons. For the most part, it’s because we want to be great investors. What I found with myself and with my team is that we really find the Kellogg Foundation’s mission to be crucial to the passion that we have for creating investment success. We don't just see ourselves as investors. We see ourselves as members of the Kellogg Foundation and champions of its mission, and that drives a genuine passion and a fire inside of us to be successful.
On the investment side, we think we may be moving into a lower-return environment, and because of that we're taking several steps in the portfolio. One step is being very focused on the absolute return potential of the hedge fund portfolio. We've spent a lot of time and have a lot of data on our funds and our managers to try to make sure that we have a true, all-weather-type hedge fund portfolio.
We're also looking at private equity from the perspective of idiosyncratic ideas. We’re targeting ideas that should generate returns in the teens with upside, but a limited downside in a market where equities decline. We think we can create this superior risk-adjusted profile, because we've always been very focused on this convexity in our portfolio. We've seen that during the big declines, we have significantly outperformed while in the up markets, mostly keeping up.
That's the convexity that I think we can really get from the combination of our private equity as well as our public equity managers who focus on value strategies. In a low-return environment, we're set to beat our benchmark and to continue to earn the real return that we need to make this fund grow perpetually on a real basis. I’m very excited and optimistic about the future of the Kellogg Foundation.