Modern Portfolio Theory in its current incarnation, involves advancements that help characterize not only investment style of investor portfolios and their components, but also there are risks. In those risks are often thought to take on multiple dimensions. What is the track record of exclusionary screens portfolio optimization tilting and the risk exposures of these ESG strategies. For some of you this might, tilt into the area of smart beta or factor oriented strategies, which also have become popular, but at the same time that ESG has risen to have such a strong position on the investment landscape. The cost of ESG when there is one maybe mineable for certain types of investing and of course it may be substantial for others, a study I did with co-authors Robert Stambaugh and David Levin, and also subsequent research quantifying the effects of ESG strategies unexpected an actual returns. Examined the cost difference as well as the risk differences of those investments on average and across various investments specifics. The cost difference may be significantly larger for equity active strategies compared to passive. Adjusting for passive risks turns out to be important. In the study, Geczy, Stambaugh, and Levin first just around the turn of the century, and then republished more recently in 2021, examined the Ex ante maximization of a Sharpe Ratio. Now that's academic jargon for trying to find the best risk-reward portfolio possible. It used expected return and volatility to measure reward and risk. The investor in this study is presumed to be one who maximizes the Sharpe Ratio, academics call that a quadratic investor because they care about not just standard deviation but perhaps variance of returns. We decided to look at mutual funds only in order to try to understand the implications for let's call it the lowest common most accessible denominator, in the investment world. It may not be only a question of whether large institutions or pension plans are sovereign wealth funds can be ESG investors effectively. But what about the little person so to speak. It allows us to define the problem narrowly, to provide a pretty specific answer, and to narrow the university even further, we focused on US domestic equity funds. In the study we asked two things first what is an investor's belief about the skill of the mutual fund manager we vary that from none to a lot, and also how good the style adjustment is. Now you've already heard about what style adjustment could mean we're going to go over that in a minute. We constructed those optimal portfolios from US equity mutual funds, for an investor who is at the standard level of a version to risk. Because mutual fund investors typically do not sell their funds short we incorporate short sales constraints and again incorporating the belief that varies across opinions on Manager skill. Now in both practice and in academics there are open questions about what is the perfect model for returns. We're going to use a number of famous models which are jargon in orientation but which are used again in practice. The first is simply to adjust for the market. In other words we're going to examine whether ESG or in our world SRI versus non-ESG or non-SRI mutual funds beat the market or do not. We're then going to adjust not just for the market on average but also the style of the mutual funds value versus growth large-cap versus small cap. Finally, we're going to augment that model with a factor that's become very popular in recent years known as investment momentum. The second model is often called the Fama-French 3-factor model, and the third is called the Carhart 4-Factor model which takes the 3-factor and has momentum. Again the three models are, simply adjusting for the market, for US domestic equity mutual funds, that's the broad market measure, broader than the S&P 500. A three-factor model adjusting for investment style value versus growth small cap versus large cap and then something called momentum which as it turns out you'll see in a minute interacts with ESG and SRI. These are called style or factor models. Finally, we've very allocations to social responsible mutual funds according to a conversation with one of Wharton's favorite sons the well-known at least in the ESG space Wayne Silby with whom we had early communications. Wayne told us that not every ESG oriented investor allocates his or her entire portfolio to ESG funds. We're going to vary those allocations, to try to understand the implications. Wayne is one of the famous co-founders of the Calvert mutual fund group. We'll calculate certainty equivalent differences, in other words, we'll calculate the simple investment utility differences between optimal ESG and total fund portfolios, and then we'll consider the correlation between ESG SRI and those non-ESG or total fund portfolios. In other words we're going to ask ultimately two questions. On a side-by-side basis does ESG underperform non-ESG, and then we're going to move to the portfolio context, and say if you are not just on a line item by line item basis but thinking about a whole portfolio, if you try to construct a portfolio from just ESG sources, is it producing a financial cost or is it not? Now for an ESG investor it may not matter in a rational world at ESG investor may indeed put up with lower diversification, or slightly higher levels of risk, or lower returns. If it ultimately produces social environmental or governance value that is important to them. Again in that sense there's no right or wrong answer. Our sample is a sample of equity mutual funds, were going to demand that at least three years of history occur into the current Manager or management team we're going to get rid of funds with investment loads which are essentially fees to come in or fees to leave. We're going to focus on one share class that has the longest data, and of course we're going to focus on those expenses. The overall sample size is about 3,500 funds about 894 after. We're going to characterize these according to 19 different dimensions of E and S and G. These were all developed from what we see in practice, what we see in the academic literature and from investment perspective says email conversations and direct contact, and conversations with mutual fund managers and their teams. First we compare them side-by-side. In Table 2 what you see is equal-weighted comparisons of ESG or what we call SRI socially responsible mutual funds in our sample, to non-social responsible mutual funds or non-ESG funds. We examined the expense ratio, which of course goes to the management of the fund. The turnover, the proportion of the investment fund that's traded per year, the total net assets which are a nominal measure of the size of the funds, the loads of the largest load share class, and of course in our subsequent analysis we issue those funds or that was still share classes that have high loads. Panel A reveal something very interesting first it looks at on the face of it that indeed ESG investments through funds here SRI mutual funds involve paying a higher expense ratio, than in their more vanilla counterparts. We find although historically higher than today, an expense ratio of 1.1 percent per year, charged by mutual fund managers to manage none socially responsible funds. Whereas socially responsible funds are about 25 basis points here specifically 26 basis points or a quarter of a percent higher on average. It's indeed more expensive to operate a socially responsible fund. It's important to understand that even for index funds, that are ESG and orientation there may be an additional layer of costs associated with the data that a fund manager needs to incorporate ESG considerations, or a team of analysts who necessarily review those data or do their own assessments. In addition, we find somewhat controverted the idea, that ESG investing is more expensive, turnover is lower for ESG funds. In other words in our sample of US active funds and also index funds, we find that a typical SRI fund has about half the turnover. It just simply doesn't trade as much. It appears to be closer to a buyer hold strategy, although by no means is it an index fund on average, 175 percent of a non-SRI mutual fund is turned over was we find SRI funds on average only turnover about 83 percent of their holdings. It looks like the historical idea of the growth of the ESG world is inherent in our numbers, non-ESG funds appear to be larger on average than ESG funds or SRI funds, slightly over half the size when comparing ESG to non-ESG. At a certain point it's worth noting that the largest mutual fund at the time in the world, the PIMCO real return fund actually developed an ESG share class. This is an example of the evolution of the industry. Finally once again a commentary on fees via the investment load essentially at a high level interpreted as a fee to enter an investment or a fee to leave, often distributed to those who market the funds. We see that SRI funds have hired loads on average than non-SRI funds. To the tune of approximately 60 basis point, s or a little more than half a percent on an annualized basis. What do we see when we do side-by-side comparisons of the average non-ESG SRI fund to the ESG SRI mutual fund world. Well panel B in Table 2 gives us the answer. There are five sets of statistics presented here, one for the market labeled MKT one for HML which is relative to a portfolio of high book-to-market versus low book-to-market firms, it's essentially the differential between value in growth in returns a style spread if you will. SMB is known as a factor portfolio which represents returns on small stocks minus the returns on big stocks, and finally momentum represents the returns on stocks that have high price or return continuation versus those that have low price continuation. These so-called factors are easily available to investors on websites like morningstar.com and other resources. They're very common today in understanding the style of investments. Finally, we offer delta some of you know that as alpha which is a measure of outperformance or underperformance on a returns basis, holding all the styles constant. For those of you who have MBAs these statistics reported are the betas or the risk exposures to the market to HML, value versus growth to SMB, small versus big, and finally high versus low momentum. The way to interpret it is as a multiplier on the differential. Now MKT which is the risk premium of the market it is exactly what typically comes out of analysis commonly done with these funds. You compare the fund to the market but you have a question about how a given fund varies with the market. For example you can see the number 0.83 in the MKT column for non-SRI funds. That means the so-called average beta when holding all these other factors constant is 0.83. All else equal when the market has beaten the risk-free security return the treasury built return, the average non-SRI fund participated in 83 percent of that return. If the market underperforms for example treasuries by 10 percent, then you expect all else equal non-SRI funds to underperform the market by 8.3 percent and so on. It's a multiplier again known as beta. In parentheses we have measures of statistical significance, those are technically speaking t-statistics. Think of it as the number of standard deviations away from zero that the statistic represents. For those of you who have gone through statistical training, you might remember the value 2 being a critical value. If you're more than two standard deviations away from zero then you're statistically significant. Putting it all together, once again we find a positive and statistically significant exposure so to speak of non-ESG funds to the market. You notice the t-statistic is 22, so it's wildly significant. What it tells us is that the average mutual fund that we consider is really associated with the overall market effect, that's not surprising at all It's essentially what you expect for equity mutual funds. However, it's interesting to note that for SRI or ESG mutual funds we find a larger beta. In other words a Beta of almost 0.9. Specifically 0.8935. Again dramatically statistically significant with a t-statistic of 24, dramatically larger than the usual critical value of 2 or could be by the way negative 2. In other words it looks like SRI funds, move with the market on average more than SRI funds do. HML represents the spread between value and growth, gross stocks are stocks today like Apple or Tesla or today once again Microsoft. Firms that have strong growth opportunities but which often don't pay heavy levels of dividends. The idea is that the market is expecting growth in the company's opportunities and the company participating in that growth is re-investing it's earnings and its profits back into those growth opportunities, and thus not paying it out to investors. The market typically is expected to pay for future growth and future cashflows that are not delivered today but are expected to grow into the future. Value stocks are different, they are typically thought to be the firms that pay out dividends currently, that don't necessarily re-invest as much in growth opportunities and which in some cases may actually represent the climbing growth opportunities or at least very very stable long-term growth opportunities. Typical stocks in the value sense are Procter and Gamble, Johnson and Johnson, although some of those business lines inside those companies may have growth characteristics as well. But on aggregate Procter and Gamble, Johnson and Johnson, and others have produced fairly stable cashflows in the form of dividends over time. Value stocks are on occasion representing older industries, heavy industries, and again growth industries are often those that are populated by technology. That technology can be today related to the Internet to hardware technology, to the knowledge economy an asset-light world whereas heavy industry is often firmly settled in value. Final word about value sometimes value can represent economic distress, a number of famous growth industries and companies turned into value companies in the area for example the land camera and Kodak, heavy steel, Bethlehem steel for example, the railroads in the US. All at one point were extreme growth areas, but of course then ultimately went through a market cycle and now look and have looked very value-oriented. What we find is a negative coefficient although not strongly statistically significantly different from zero for both SRI, non-SRI or ESG and non-ESG. The non-ESG non-SRI mutual funds, appear to have a negative coefficient of negative 0.036, whereas ESG investments appear here in mutual fund form to have about a negative 0.03 level. Very close and statistically indistinguishable. What does the negative sign mean, it means that the average mutual fund in our sample appears more growth-oriented than value-oriented, and that's an historical artifact that we see very strongly today. Mutual funds are often investing in growth as part of the preference of investors. Turning to small minus big, we start to see once again statistically significant important economic differences. For example, historically it looks like ESG investing for the so-called common person through mutual funds with typically low initial investment levels has a 0.16 exposure to the SMB effect. Broadly wildly statistically significant with a t-statistic of 9.7, compared to the ESG SRI estimate of point 20. Again with a very strong statistical significance with a t-statistic at 12. Again this is a technical conversation, but the intuition is that it looks like ESG investments have tended to be smaller historically, the non-ESG investments looked at via the holdings of mutual funds. Now today of course in the comparison for example of ExxonMobil, versus what is sometimes a largest company on Earth Apple. The size, comparison might move the opposite direction, but in general investing in ESG has been in part investment in smaller firms especially with growth characteristics than larger firms often with value characteristics. Finally, the largest difference we see is in momentum. Momentum is a characteristic of pressure in prices or movement in returns across time. It's a very strong effect and if you haven't heard about it I encourage you to do some additional reading on momentum. It's a kind of as of yet less than perfectly explained effect in the academy, but it's very important for selecting investments and it's very important across asset classes. It's mentioned in professional circles all the time and again if you look at websites or or services like Morningstar or many others you will see this momentum characteristic mentioned in describing investment style and investment risk. When an investment has a higher exposure to this momentum effect it means there's a price pressure or movement in the returns for the investment. It's associated with higher average returns. What we see here is that ESG style investments as typified by SRI mutual funds have about twice the so-called exposure to the momentum effect as more traditional investments. You can see it here 0.05 versus point 0.28, both statistically significant. What's the implication, at the point estimate? It says that being involved in ESG investing historically has involved a momentum effect in returns. By the way, momentum is often the thing you see in venture capital when we can measure it or certain technology investing, which may make sense. We also see differences in growth and value among those classes. It's a pretty gear headed technologically enabled analysis that we just went through the final question for a fiduciary or really for any investor who's trying to isolate at the lowest common denominator level whether ESG can live with returns whether good and gold can live together is to ask what happens on all of that is held constant. Momentum, firm size in the investments of the funds, HML, and general market exposure, all are held constant? What do you see with net returns? Well, the answer is in the first column, here labeled Delta the outperformance or could be underperformance. While we don't find strong statistical differences in either cases, which means to us that the model actually absorbs substantial amounts of mutual fund returns in general, we do find that at least at what we call the point estimate 0.0021 versus 0.008. ESG actually outperforms non-ESG. What it seems to say, at least is that ESG once again appears not to do harm. This is highly consistent with the more recent investment analysis that you've seen and on a side-by-side basis which is often inherent in the guidance from regulators it looks like ESG has not provided a cost on a side-by-side basis. However, the question still remains when we move to portfolio construction do we pay a cost associated with lack of diversification inherent in the focus on ESG? Side-by-side comparisons, while favored by regulators do not give us the full answer. We have to turn to portfolio construction next. One of the most pervasive measures of fund historical performance, is known as the morning star ratings. Those ratings range from 1-5,one being low five being high, and tell a story with historical I, as to whether mutual funds have performed well or not. It's a little bit like restaurant ratings perhaps, that are falling on a 1-5 scale. Around our university when people are on campus for graduation, typically we see people going into four or five-star restaurants if they have a chance for graduation dinners. It's rare that we seek one-star restaurants although that can be fun as well. When we looked at an historical snapshot for the data we considered in the study, we found that SRI funds and non-SRI funds have an interesting pattern which is at SRI funds historically have had higher morning star ratings than the entire universe put together. This is data from the US Social Investment Forum Foundation famously known as US SIF and we think that this continues into the most recent time period. For our study what is it points out is that our results are reflected in common measures of fun performance. As any investor or practitioner of Modern Portfolio Theory, knows portfolio construction really is the key to investing at the end of the day. What we tried to do then is according to different investor potential believes create portfolios based on capital market expectations inherent in the technique that provide the greatest ratio of reward to risk or the highest Sharpe Ratio. This again is said in the context of what we call retail investors, are those investors who might have ability only to put small amounts of money into play here into mutual funds. The nine categories you see on the table incorporate first your belief in a model or an approach to investing, here it's going to be index investing and that's called the capital asset pricing model has been made famous in academics by Bill Sharp, who won a Nobel Prize for the work in 1990 along with Harry Markowitz and others. But then we're also going to allow within each of those categories about whether you think index investing is right or wrong, to invoke the idea of manager skill. That's important today because while there have been trillions of dollars flowing into passive investments, and the famous founder of Vanguard Jack Bogle and others would tell you that managers don't on average have skill and that you should be a market indexer. There's still a lot of money going into active investments, and some might say that ESG investing by itself is active investment. Here we're also focusing on the ability perhaps to beat the market, so that gives us nine categories. Three categories of how good index investing is, and then three-quarters within those about whether an individual or a group of managers have investments skill. For those of you who want to delve into the statistics we're giving you the link to the paper, which recently was republished in the review of asset pricing studies. The intuition is that there's an increasing probability of index investing being wrong as we go from left to right. You can think of it as a high probability say, 60 percent of the model being off by half percent per year for the middle column or 60 percent probability that the model is off by one percent in the far right group of three columns. Skill uncertainty, that's just saying again whether you believe mutual fund managers can beat the market. In the first column we're saying there is no probability, in the second column of each group of three, we're saying that there's 30 percent probability that a manager net of fees can beat the market by half percent and in the far right column under each group of three were so greater skill we're saying that there's 30 percent probability under manager can deliver 1.5 percent of outperformance. Now this is just a technical calibration from the bell curve. The key realization in the table, is that as you look down the column that is labeled 00, you understand first that this is a market indexing column. That says that the market index approach the so-called Capital Asset Pricing Model or CAPM is right there's no probability it's wrong and there's no probability that a manager can generate value-added from skill. This is what I would think, Jack Bogle the founder of Vanguard was thinking when he gave the advice that everyone should just passively index their investments and let the market do its job. What do we see? First index oriented funds are chosen among the ESG optimal portfolio components. We see for example the dominant social equity fund which is a fund related to the KLD index that we mentioned earlier. We see an institutional share class for a fund called the Lutheran Brotherhood. We see Liberty Young Investor Funds K class it's also a cheap institutional share class. Then we see the AHA Investment Fund a Diversified Equity Portfolio. Now if you were to invest in the broader universe non-ESG or ESG plus non-ESG, you might actually see investments in Vanguard low-cost mutual funds or those offered by other institutional investors BlackRock and Vasco and others. Here in the ESG universe, the model is trying to find their counterparts and so it's trying to replicate the market portfolio. You can do it, potentially with these five investments according to our results. What's more than that if you look at the correlation between the best fund overall and just this ESG subset, you find that it rises to the correlation level of 98 percent and you only pay a cost of seven basis points. It's a pretty strong result that suggests, that ESG investing while it might involve some measure of cost it's certainly not very large in the eyes of some investors. That is to say if your market indexer there are available historical options for you. However, as you start disbelieving a bit in index investing according to our definition, or you start believing that managers have skill, you suddenly start paying costs. In addition your portfolio rotates for instance if you go from believing still that you should be a market indexer, but generally speaking there is some skill to be had you shift your weight increasingly to the diversified equity portfolio and you start adding wealth to the barren growth fund. Historically high performing fund focused very heavily on growth which again, is not that surprising in the ESG universe. However, you're giving something up. The number of investments goes from five now to three, and it turns out that the correlations between the best ESG portfolio and the best total portfolio go down to 94 percent. If you go all the way to try and drain a huge amounts of skill, you move into the high skill fund, Barren growth fund and you also diversify it with a mid-cap fund and so on, but it looks like the correlations go down on the basis of the smaller number of funds in the adherence to ESG. A similar feature is in place as you go from left to right in the overall table, so it's a fairly technical discussion but I think if you give it a moment of thought, you realize this is logical. While a side-by-side comparison between ESG and non-ESG mutual funds for the average investor and accessible to many, may not produce much of a difference. When we dramatically across 19 different styles of ESG consideration, limit the investment universe you start paying a cost. Again I highlight that this may not necessarily mean that it's wrong for investors to pursue if their own utility is improved through having the characteristics of screening out investments. But it also goes to show that if the constraint binds there might be a legitimate costs to think about. Another way of saying it is that if we get rid of every investment that is anathema to somebody, there's actually relatively little to invest in. It involves a significant measure of concentration. What have we shown, in this examination of an attempt to create optimal fund portfolios for the common investor using the historical data? Well first on an aggregate basis, the ESG SRI constraint does not appear to tightly bind for a typical portfolio for those mutual fund investors seeking the highest risk-reward trade-off. In addition, in the case that's specific to market index investors, the expected ESG cost appears to be low for reasonable parameters and for reasonable believes. However, in our results and other results that are in our research this constraint associated with ESG exclusion, has expected costs. That's especially true when one focuses on style in part because of the extreme style bias that we have in some cases. Or when we believe fund managers have skill, while it's possible to find high-skilled ESG managers, the numbers are smaller than in the broader universe. It ends up being simple math. If you have a binding constraint, if we have investments under exclusion that leave out certain risks or rewards inherent in the investments that might be valuable for investors there might actually be a cost. It ultimately points out that the examination of those costs is critical. While there's an ongoing movement perhaps today, to include ESG investments and retirement portfolios it's something we should give careful thought to. Personally I'm an advocate of it, in part because I think investors should be creating and investing according to their own preferences, which legitimately move beyond only a risk and reward. But for those who managed the money of others, it requires specific assessment in my view and I think the evidence supports it.
