Tue, Nov 12, 2019
A A A
Welcome Guest
Free Trial RSS
Get FREE trial access to our award winning publications
Horizons: Family Office & Investor Magazine

Peter Kellner: Asset allocation and machine learning lift ESG multi asset portfolio to massive outperformance

Tuesday, March 05, 2019

After a 3 year data science build creating a deep competitive edge in the use of non-financial metrics, Peter Kellner and his all-star team managed to combine two of the most critical megatrends of the 21st Century : Integrated Sustainable Investing & Artificial Intelligence in a multi-asset strategy which is up net 30.7% since June 2017 inception. In this feature we are looking what’s behind this remarkable success.

Matthias Knab: As a start, Peter, tell us a bit of your background and you got started with ESG?

Peter Kellner: I first got involved with ESG and sustainability through my lifelong mentor, a man named Bill Drayton. He is the founder of Ashoka which is the largest organization globally of social entrepreneurs. I met Bill when I was at Princeton in college. Bill has been responsible for creating some of the most impactful social organizations around the world, like Teach for America and City Year. He inspired the founders, and he inspired me to try to live a purposeful life. I knew I always wanted to be in finance but I also knew following him that I wanted affect change.

And so, right after college I went to Hungary, to Central Europe which is where my family originally is from. I was on a Fulbright scholarship and I was struck by the state of the environment and the way the Soviets had left Hungary, which was very devastated. I set up an environmental policy organization there that I modeled on the Natural Resource Defense Council, which is an organization that I followed when I was a teenager in the United States, a very large and important organization. I was able to bring together Hungary’s top scientists, journalists, economists, lawyers to do interdisciplinary reform of the environment and the transition to a market economy. We actually ended up as an organization writing the environmental evaluation for the approximation program for Hungary to ascend to the European Union.

Today, about 26 years later my organization is perhaps the most influential organization in Central Europe, training judges, prosecutors, writing laws. It also has become global, it works with the World Resources Institute and others, and I am still involved and contribute to it. That was my major first involvement in sustainability.

My second one was also guided by Bill Drayton. I traveled with Bill to many places after college, specifically Latin America. I developed an interest in Ashoka’s model to identify social entrepreneurs who would have a high impact. I loved that idea because I was also business-minded, and it was clear to me from my work at undergraduate and just living abroad as I have been doing in Central Europe and Russia, that entrepreneurs were going to be solving the problems in the emerging world.

If you think back to the ’90s, the emerging world was truly emerging then, with massive growth of the middle class across the board, and a lot of it was driven obviously by the internet and the surge in technology. I co-founded an organization called Endeavor with a woman who at the time was working for Bill Drayton at Ashoka. I spent years building that organization on the ground in Latin America and then I helped expand it into the Middle East and ultimately into the United States. My agreement with my co-founder was that I would go onto a career in finance and entrepreneurship but stay involved with the organization at the board level. The organization is now in 33 countries around the world and doing impressive work in terms of the entrepreneurs and supports. Last year, those entrepreneurs generated $16 billion in revenues and has created over a million jobs.

And so indeed, I did turn to finance, and my day job for the last two decades has been Venture Capital and entrepreneurship. I’ve built some companies and have also been an investor in some wonderful companies in the early stages, companies like Salesforce, Learnvest, AdChina, Dataminr and others. I’ve always loved to work with entrepreneurs. My specialty is in financial services and and data, and their nexus. In fact, for the last fifteen years I’ve been focused on data science, beginning at a time when data science and machine learning weren’t the headline every day.

Because of my passion for sustainability, I had the idea many years ago to see if I could figure out a methodology to use data science to identify what we call material environmental, social and governance (ESG) metrics that you could then apply to asset classes to generate Alpha and also minimize risks and liabilities when you were doing your financial analysis. This was public securities because my world has been private securities. I explored the world of sustainability and investing today, it’s largely private, 95% of it is private.

If you look at the world of active public investment in sustainability, it’s only 4% - 5% of the $180 trillion global liquid capital markets and that’s net of derivatives and OTC, another $700 trillion plus. There’s this disequilibrium where you can potentially have a massive leverage in the public markets but it’s barely penetrated. I wanted to figure out a strategy to penetrate those very large markets with the most impactful strategy I could think of: using machine learning – which in the end is data science – to identify ESG metrics.

Here it is important to understand that ESG metrics as they are today, in early 2019, are terrible. They are self-reported by companies, with many vanity metrics. They are non-periodic – the time series is generally not particularly well put together, and it’s not well-weighted. The metrics providers we have today don’t have particularly good information. If someone is actually saying they are using ESG metrics for alpha creation in their investments, that’s not a credible claim unless – and this was my hypothesis which proved true since we started trading – you are actually using data science to identify those material ESG metrics that can be non- financial and sometimes financial metrics that affect financial performance.

If a corporation does a 10 kilometer run on a weekend for its employees, they may contribute that toward their ESG metrics. However, that’s not really a material ESG impact. It may be good for the brand in some minor way. But if the same company - say it’s a bank or a real estate holding – has 10% of its US real estate portfolio in recurring flood zones, well that’s a material ESG metric.

But most financial analysis is not looking at such particular metrics and off-balance sheet items. However, those factors are incredibly important in the generation of Alpha which has been proven out academically by George Serafeim of Harvard Business School, who is also on our board and advisors at our fund. Professor Serafeim showed that if you use a materiality lens when looking at ESG, you can generate more than 6% in additional Alpha. This approach also minimizes your downside because you’ll be looking at liabilities or regulatory concerns which otherwise may not get identified.

With my approach, I brought a couple of different things together. I looked at the disequilibrium of market size and the ESG adoption of both private and public sustainable initiatives and focused on the public securities as the domain where we need to engage ourselves now if we are going to have any chance of turning back the tide of climate change. And in doing so, what is the most interesting, largest sizable strategy I can pursue?

All strategies today pursuing ESG in the public markets are predominantly long-only equity, maybe with a little credit. I went out and did on my own data science study in 2014 and to 2015 looking at securities in Western Europe, the United States, as well as sovereign securities, to see if we could use machine learning identify Alpha. The results were stunning in terms of the p-values, the correlations, the sharp ratios, etc.. We were able to show that one could find material ESG both in corporate as well as sovereign categories. This got me terribly excited.

So, I founded a firm to trade a multi-asset strategy. What’s exciting about this approach is that such as any active manager, we want to deliver superior performance, which we have been doing. However, we do that through the integration of material ESG metrics, and so we together with our investors are accomplishing two ends. First, we are not sacrificing performance to get to a contribution to planetary sustainability. We are actually using the very same mechanisms that you need to ensure planetary sustainability, to generate superior performance. It’s a beautiful business model. We are also investing in the gigantic public markets with a strategy – multi-asset – that has no limit on your assets under management. The only limitation you have is your rate of growth. If you’re a long-only strategy and there some wonderful ones out there, you’re actually going to be limited in how sizeable you can be because markets are only so deep for those strategies.

We look across the world. We look at every asset class. We’re able through our machine learning platform that we spent a lot of time developing – a substantial proprietary platform that both cleans data sets that are out there, generates new data sets, new algorithms. We use that to provide our traders’ information to be more intelligent when they are making a decision about an asset, and how ESG -- material ESG relates to that asset. We’re not a quant fund and we don’t want to be confused for one. A quant fund is a very different animal. We are using quantitative technologies to augment our knowledge of trading, so human beings can actually carry out those functions. I don’t really think that, today, there’s a way you could ever use 100% machine learning or AI to find ESG, because at least 20% of ESG metrics are very qualitative and they require that kind of sourcing.

If have to put a ribbon around it, you have the recognition that if we’re going to make real change, we have to take on the global capital markets and we have to do so in a way where we can attract investors to the idea through superior performance. We can achieve that by using the methodologies of integrated ESG investing, because by now it has been proven that material ESG metrics generate Alpha and limit your liabilities.

Matthias Knab: Peter, this fascinating because I see that with your methodology, you’re actually addressing some of the structural and procedural problems that ESG has at the moment. Can you tell us more about your process, and where would you draw the delineation line to impact investing?

Peter Kellner: Let me point out first that for me, impact investing is an asset class whereas integrating ESG into your financial analysis is a methodology which can create Alpha and mitigate your liabilities. They are entirely separate. Our strategy is not impact investing, which as you know runs a spectrum from purely philanthropic to seeking market returns, but always with a social end in mind. We are different in that we actually use the benefits of contributing to planetary sustainability through identifying these key ESG metrics that go into our financial analysis that produce superior returns.

It is a very different proposition wherein we’re not sacrificing in any way return for social good. We are augmenting return and social good. But what we do is not within an asset class. Structural challenges that exist today have to do primarily with the research with the information available, which is not standardized – like I said it’s non-periodic, it’s self-reported, it’s in general very bad.

That’s one of the primary issues we have. There are also people who are very much in disagreement over what are the actual material metrics for a given asset class, so there is a lot of disagreement which really is based around the lack of standardization. That is something that has to evolve in the market.

We feel we addressed these two issues through the development of our proprietary technology, because what it allows us to do is to find that needle—in— the—haystack, to find those typically four to six material ESG metrics that apply to a corporate or sovereign asset. In doing so, we can be consistent, so we can contribute to the process of standardization across the industry. We can identity a line of sight from A to B that our identification in that particular metric produces that Alpha and that positive change. It’s highly measurable, and measurement is another structural challenge across all impact investing and integrated ESG investment. We are addressing this through machine learning and through also by the combination of our technology with our team.

I should also make a point here about the strength of our team. They are fantastic, every member of our team is certified in the FSA which is the sustainability accreditation that’s equivalent to the CFA. Every member of our team has taken the Level I & II FSA, which is under SASB, which was started by Michael Bloomberg. SASB is like FASB but for the sustainability sector. All our team members have deep knowledge of ESG through having gone through this accreditation. It’s not like we have our environmental people over here, who may have studied environmental science but have no financial background and they integrate with our trading team over here who are really excellent traders but don’t really understand ESG and they come together and figure things out. No. We have an integrated team that is profoundly good at what they do in financial analysis and investing, but they also have very deep backgrounds in ESG.

With our strategy, we find ourselves highly differentiated along many aspects of the ESG spectrum and fund management. For one, we are to our knowledge the first to create a global ESG multi-asset fund. The second is that we are using proprietary machine learning that we have developed with a great team of data scientists, to identify material ESG factors that affect financial performance, drive Alpha and minimize liabilities. We are also using a hybrid process where we combine exceptionally qualified teams with our technology. That is something that very few firms are doing.

These are the things that I believe differentiate us from other funds right now, and frankly, I encourage other funds to get into our business because that’s how we are going to change the world. The other aspect why we don’t really fear competition is that we believe we have already built a relatively wide moat in terms of our understanding of sustainability and the appropriate metrics together with our proprietary technology and its capabilities.

We believe that we have very differentiated products and strategy, but apart from the superior returns that investors can enjoy, it is also very needed. And here, our appeal is not to your heartstrings, but really to your rationality, Sustainability is the most important mega trend of the 21st century – we are facing major challenges and continue heading toward a cliff. How are we going to address that? My view is that the capital markets are the solution. As I tried to explain, throughout my life I have had a history of believing that capitalism can drive massive change, and I mentioned the Endeavor organization as example which I co-founded.

In the same way, I believe capital markets are likely giving us our highest probability of a course correction in environmental degradation. That’s not to say governments and NGOs are not important – they are very important. But capital markets have not been tapped yet to do this work and yet they are so sizable and so influential. The capital markets possess an enormous palette of tools with which to invest across the world in different nations, in different sectors and categories.

This is important to mention as I believe we are never going to run out of ideas. We know that in the world of hedge funds and finance, investment strategies are usually limited and that Alpha can decay very quickly. People run out of ideas, and often close their funds or if they made enough money, convert into a family office. I think that this ESG strategy is a gift. There will always be numerable ideas and strategies you can pursue, whether it’s green bonds, real estate, or in better supply chain, in different asset classes. You can apply this to currencies just as you can apply this to equities. It’s truly amazing.

Despite all the feel-good aspects of it, it’s really critically important to understand that this ESG strategy, at least how we understand and run it, is not based in some philosophy or moralistic point of view, but on empirical research. Professor Serafeim from Harvard Business School, who is on our advisory board, is the world’s leading ESG analytics professor who already in 2015 authored a paper published called the “First Evidence In Materiality”, where, as I noted, he demonstrated statistically that you can increase Alpha by focusing on material ESG in your corporate security selection. This is empirical and I put a lot of resource into my own study where we worked with a very well-known data science firm for over 12 months to see that this was applicable, not just a corporate equities and debt, but also to sovereign asset classes.

This is an empirically based strategy – it works, it’s credible, and I believe our returns reflect that. Since inception 19 months ago, in June 2017, we are +28.32% net, having finished 2018 +9.99% net.

Matthias Knab: Looking forward, how do you see the ESG space further develop?

Peter Kellner: We believe that within about 15 years, the integration of ESG into financial analysis is going to be normal. Today, it’s not and there’s a lot of resistance to it, there’s a lot of short-termism in the market place amongst CEOs that look at quarterly earnings calls, there’s a lot of pressure on the very things that prevent an ability to have a longer bias. But when you are trying to invest in sustainable assets and apply a sustainability lens, you need a longer period to accomplish that – you need three to five years to start implementing your strategies and also until the companies you invested in get results from the business plans that they have created around sustainability.

We mentioned that in the world of hedge funds it is kind of standard for managers to pile into ideas that over time those returns to be arbitrage away as everybody piles in.

In this particular area, the opportunity for return seems to me to be infinite because the possibilities for creative investment are infinite. We have really just begun, and the more people pile in, the better we are all because it validates the space. Still a lot of people, particularly in the US versus Europe, question ESG and its relevance. Professor Serafeim of Harvard Business School did a study on this looking at asset managers in the US and Europe, and he demonstrated that in Europe asset managers recognized ESG as something that both enhances return and mitigates risk. When they get asked the question, “Do you contemplate ESG?”, they say, “Why are you asking the question? Of course, we do!”

In the US, the managers in his study, who were managing 10s of billions of AUM, tended to say, “In large measure, we approach ESG when our limited partners asked us to do so,” but they themselves don’t really tend to recognize the intrinsic value.

Matthias Knab: Tell me more about your portfolio construction process.

Peter Kellner: Portfolio construction is actually designed by my partner and CIO, Decio Nascimento. Decio is remarkable, allow me to introduce him here with a few words. We met completely fortuitously through a mutual friend at Paulson. Decio had started his career at Goldman Sachs, trading derivatives when he was putting himself through The London School of Economics. After that he went to Brazil, he is Brazilian originally and worked for Luis Stuhlberger at the Verde Fund – one of the most successful macro funds in the world.

Matthias Knab:Bloomberg called Luis Stuhlberger “Brazil’s 15,000% hedge fund legend”. He is generally not that well known, however I have been aware of him for a long time as I have been to Brazil a number of times and dealt with hedge funds there.

Peter Kellner: Right, and so you would probably like to hear that Decio was Luis’ right-hand person 24/7, and so that was where he learned his trade and Luis’ philosophy and his approach to investing. Decio then wanted to move to New York, however Luis did not at that time want to set up an office here and so Decio joined 3G Capital, which is a well-known and successful firm who principals own by Anheuser- Busch. They acquired Heinz and Kraft and Hortons and others, and are very successful investors.

At 3G Capital, Decio helped to run their liquid strategy and was responsible for macro. Decio then took a sabbatical because he had literally been working since he was 18. We met just as to New York, exactly when I had finished that proprietary data science study that I talked about that corroborated my thesis about materiality in ESG metrics as applied to corporate and sovereign asset classes.

At that time, Decio had offers from significant macro funds, but he wanted to do something impactful. He told me, “In my time off, I have been thinking about it, but I couldn’t figure out a way to use my skills to do something impactful,” and I said “Well, here’s an idea, why don’t you join me and we can take your deep skills in multi-asset fund management, and combine them with ESG.”

We have an institutional scale back office today, focusing on risk, compliance, cyber security – you name it. Decio took over the portfolio construction and I built and manage the team strategy, including our data science. Decio was also part of the first cohort of the The Sustainability Accounting Standards Board’s Fundamentals of Sustainability Accounting (FSA) credential, so he was in the first class to qualify and he made sure that everybody we hired thereafter in the firm got that accreditation. More than that, he went and took the MBA in sustainable finance at Yale, which is taught partially by one of our advisor board members, Dan Esty, who was also my professor at Yale Law School. Dan is one of the foremost well-known professors in environmental law in the world.

Decio and I began to build that machine learning platform from the get-go, together with Erik Allen who is our AI director, and advisory board director, an MIT PhD in Artificial Intelligence. We did it through phases with a team of seven full-time data scientists. In the first two phases, we were able to do all the scrubbing of datasets and then the creation of new data and all the interpolation to identify those critical material ESG factors.

We are now in that third phase and developing what we have called the Compass Materiality Dashboard. This is to my knowledge the first dashboard of its kind with respect to ESG metrics and publicly trades corporate and sovereign securities. It will enable us to identify each of those four to six key material ESG metrics – E, S and G for every asset class that we look at. I personally think that this is stunning breakthrough in investment and market intelligence.

Decio is in charge of portfolio construction, which is all about asset allocation. His process is a top-down, traditional approach where we are looking at the big blocks, ECB, BOJ, China, the Fed, and their interest rate policies which affect everybody else. Then we drill down from nations to sectors and then down to specific companies.

At every level of our analysis from the macro to the idiosyncratic, we are also integrating ESG. When we are looking at sovereigns for example, beyond the macroeconomic work that we do to establish what’s going to happen to the world, are there going to be rate hikes, et cetera, we are also looking at what we call “the health and wealth of nations.” For that, we are using indices that are non-GDP focused. Of course, we do use GDP-focused, traditional measurement, but also non-traditional GDP indices. One important one that we’re using is called the Social Progress Imperative founded by our colleague, Michael Green, and our fund has a formal collaboration with SPI.

SPI has developed these comprehensive indexes across the world, measuring nations and down to the municipality level in different dimensions such as children’s access to education, primary, secondary, higher educational access too, corruption levels, women’s rights, environmental degradation. The list goes on and on and on, and while these are not standard GDP measures, they give you a view of the health of a particular nation.

If you combine that with the traditional monetary indicators, you then can get a rich picture of a country when looking at it from the ESG perspective.

Apart from this traditional top-down construction we also have a more technical management of the cash book and the trade book that we engage in. The cash book for example, we’ll have an expected yield, say 10%, and that’s going to be dividends in capital appreciation, equities and debt. The trade book will be derivatives. We’ll have our expected yield in the cash book and that way, we’ll size the trade book. Let’s say, at 6-8%. In the trade book, we will put on asymmetric trades 3 to 1, 5 to 1, 10 to 1, 20 to 1. Based on the performance of those trades, we can have very large pay outs if we just hit two or three of them. It’s also a way for us to invest without having too much cash exposure.

As investors, we have a long-term horizon and aim to have a low turn over in our portfolio. When Decio and I came together in 2016, I remember him saying to me, “I anticipate four rate hikes, by the end of 2018 and a meltdown in the emerging markets. I think we should position the portfolio for that,” which is precisely what played out, by the way. So, we were well-protected against some of the volatility that happened end of 2018 and were able to end the year with a b net return.


 
Today's Exclusives Today's Other Voices More Exclusives
Previous Opalesque Exclusives                                  
More Other Voices
Previous Other Voices                                               
Access Alternative Market Briefing


  • Top Forwarded
  • Top Tracked
  • Top Searched
  1. Mexican infrastructure funding helps alternatives investors[more]

    Laxman Pai, Opalesque Asia: Certificates of Capital Development (CKDs) are helping investors to access a variety of developing sectors in Mexico - namely infrastructure, real estate, mining, technology, and business, said Preqin. Introduced by the Mexican Government in 2009, CKDs are private c

  2. Tech: Competition heats up among technology providers like Backstop Solutions, Solovis, and Northern Trust, Don't forget the human touch in tech push[more]

    Competition heats up among technology providers like Backstop Solutions, Solovis, and Northern Trust From Institutional Investor: The era of Excel spreadsheets and manual data entry are over for institutional allocators - or they should be. Another technology product just hit the market,

  3. Fund managers face crackdown after Neil Woodford scandal, More than 500 Hargreaves clients join Woodford lawsuit[more]

    Fund managers face crackdown after Neil Woodford scandal From The Times: Asset management giants have been warned that their funds could be suspended and wound up after just three months if they do not ditch riskier stocks in the wake of Neil Woodford's downfall. The threat has been

  4. Ackman, down in October, has a new position[more]

    From Institutional Investor: Bill Ackman isn't done with shaking up his portfolio this year. After trimming several of his biggest positions that soared this year, Ackman is building a new undisclosed stake that already comprises 8 percent to 9.5 percent of his portfolios, he told investors. P

  5. PE/VC: Private equity is the new stock, Private equity fundraising in the US hits all-time high, Foreign private equity firms lead $152bn blitz on London Stock Exchange[more]

    Private equity is the new stock From Institutional Investor: The traditional portfolio of stocks and bonds needs an alternative investment shake-up after failed monetary policy, according to executives at JPMorgan Chase & Co.'s asset management unit. "We are in an odd cyclical posit