Are University Endowments Too Risk Averse Now?

University Endowments

The diversification strategies that what is called modern portfolio theory aren’t so modern anymore, and after almost 70 years, are in need of modernization again.

Modern portfolio theory came out all the way back in 1952, in a paper by economist Harry Markowitz. Markowitz ended up winning the Nobel Prize in Economics for this feat and this theory remains at the forefront all these years later.

This approach to investing did modernize the way we looked at the subject, and essentially provides modeling for how we may maximize our returns by way of diversification given a constant risk tolerance.

Like all economic models, there are certain assumptions that need to be built into the model, so the task then becomes one of discovery with these assumptions in place. If this model did maximize returns with these assumptions in place, then it would succeed on that basis.

Modern portfolio theory also makes another big assumption which is that assets are to be held, and then we look at what asset mix would provide the best returns based upon our assumed tolerance. Provided that we buy in to this particular investing strategy, then investing according to this theory would have us pointed in the right direction.

Modern portfolio theory does work under these constraints, but the obvious question then becomes how optimal this style of investing is. In a real sense, modern portfolio theory locks us in a closet and then tells us the best way to invest while stuck in it, and while they may do a good job at this, the bigger question is whether being in this closet is the best idea or even a good one.

Modern portfolio theory still plays a big role in institutional investing to this day, and in particular, with large university endowments. When we see them miss the mark like they do, some may question their investing strategies, but they always have the excuse of using underperformance as the price of the safety that they aspire to.

The idea that risk and returns are opposed comes from the idea that we need to approach investing statically, meaning that we choose a certain asset mix and certain assets within them and hold them while looking to balance the risk between them.

What this serves to do is to divert the attention away from the risks of the investments themselves on a temporal basis and instead usse long-term analysis to assign a certain level of risk and then look to allocate assets according to these measures.

Getting the Right Asset Mix is Important, But the Calculations Must be Relevant

A simple example of this would be with the way that a lot of investors use asset allocation to manage their own portfolios. You look to allocate your investments across two or more asset classes, stocks and bonds for instance, and then seek out what the best allocation would be to keep risk within your tolerance level and seek the most return.

We don’t really do much calculating when we do this for ourselves, and might just use a chart or a vague sense of our risk tolerance and assign the asset mix accordingly, but you can also do some math here and get a better idea of what the optimal mix is. This is what modern portfolio theory does in a nutshell, and it certainly is better than just making a guess at this like individual investors deciding this on their own or with the help of a salesman usually do.

Modern portfolio theory is therefore better than guessing, and when you’re managing a $40 billion dollar endowment fund like Harvard does, the wrong guess can be pretty expensive.

The problem with this theory is that investment risks are not constant, meaning that the risk at one point in time is going to be different and often quite different than the risk at another point in time. We need only look to how the stock market moves to understand this completely, where risks clearly vary and the need to hedge them varies as well.

In times when the risk is lower, this will have this theory underperforming the market, in other words have them over hedged. When the risks are higher, since this hedge is designed to address an average amount of risk and does not really adjust, it will leave us under hedged.

In both cases, the theory has us underperforming our potential, where in one case we’re limiting our risk too much, and in the other, we’re not limiting it enough.

Since we’re talking about managing asset allocations, this is not a matter of being in the right or wrong investments within the asset class, like for instance being in the right stocks, as while this matters as well, that’s a separate issue from looking to optimize your asset mix, which is what portfolio management seeks to do. Before we can decide what stocks to be in, we need to first decide how much of our portfolio that we want to have in stocks, so this is an even more fundamental matter.

Since stocks have the highest potential for return, we want to be in stocks to the degree that their risk to return ratio is favorable, and more favorable in fact than other options such as bonds, real estate, private equity, or whatever else we’re looking to invest in.

Modern portfolio theory actually gives more weight to risk management than to seeking returns, which is a mistake if we’re actually seeking an optimal balance between them, because this will be shooting for a sub-optimal balance. We need to find out what the optimal balance is first, and then we can adjust up or down depending on whether we want or need to give more weighting to risk control.

An institution such as Harvard should not be risk averse and should be neutral. We need good reasons to be risk averse and there is no good reason why this endowment would need to over hedge, given its structure.

That’s not how portfolio managers of these endowments tend to see things, although their perspective is shaped more by the status quo than anything, where it’s seen to be safer to just do what others are doing here and reduce the risk that you will be blamed if things don’t go well. If you just follow the crowd, then you can all blame the markets.

The numbers for Harvard’s endowment fund just came in and they realized a 6.9% return over the past 12 months. People are pointing to the 10% gain that the S&P 500 gained during this time, looking to cast a shadow on the fund.

Managing Risk is Very Important, But Risk is Far from Constant with Investments

The fund will argue that they can’t just look at one side of the equation, the return side, without seeking to manage the risk side, and they therefore can’t just expose themselves to an allocation of 100% stocks and buy indexes such as the S&P because if they just did this all the time, there would be no risk management going on at all.

While this is true, this does not have to be an all-or-nothing choice, or a decision that needs to be made once or only occasionally, and that’s the big mistake that they make with this reasoning. They want to choose an asset mix that works in all seasons, but seasons change.

It doesn’t really make sense to choose a form of dress that will minimize our discomfort year -round, like wearing a sweater. We’ll be too hot in the summer and too cold in the winter but since we can only choose one method of dressing year-round, that may be the best choice, if we had to that is. The fact that we do not somehow gets lost.

It’s summer now, and they have their sweater on, so this just isn’t going to be optimal for the weather. There is a view out there that we can’t really predict the weather with investments, even though we somehow can predict good weather in the long term, but changing weather patterns along the way are somehow random and unpredictable.

In reality, the shorter-term forecasts are more reliable than the long-term ones, just like they are with real weather forecasts. While we may use complex models, it’s actually not difficult at all to get a sense of what the weather is like, you can just go outside and observe for instance.

With the stock market rising, this should naturally skew us toward seeking out more the returns that this is providing, because this provides us with a favorable risk to return ratio overall. Markets that are rising are less likely to drop, and markets that are rising also provides us with the more generous returns that stocks provide. In this case, while we may not want to put all of our eggs in one basket, we should want to put more of them in this one, and less in it when the risk to return ratio turns sour.

When the stock market is falling, risk goes up and expected return actually turns negative. This is the time when we really need the hedge, and we could even say that a model that just relies on stock market performance to determine asset allocation as far as what we commit to stocks would at least be looking at the right thing, with the only other consideration being what other assets we will be using in our mix.

Being over hedged during the good times can make people look smart when the bad times come, as you will beat the market during this time by having a certain amount of your portfolio invested in assets with better performance. That’s definitely a good thing, but it makes much more sense to use this tool as it is actually needed, and to also use it in much more abundance at times than a static allocation could ever provide.

Being fully invested in stocks is actually not that bad of an idea for an endowment fund at the right time, as is being fully divested in them is when they are performing poorly. They could be excused for not going all in, as with funds this big it will take some real time to re-allocate, but we should be striving or not an optimal level for all seasons, but one for the current season.

There’s a lot on the line with $40 billion in play, and should we suffer a big crash like we did in 2008, you don’t want to be losing $30 billion’ of this even though these losses may just be mark-to-market ones on paper. If our strategy has us under hedging against such things, our losses with a static strategy are going to be bigger, as well as our gains being smaller when we do gain.

Harvard’s hedge over the last 12 months came at a cost, and this is a cost that is easy to figure out because we can just multiply $40 billion by 3.1% and get $1.4 billion. Just knowing how these funds operate, and in particular, the fact that they do not seek to hedge according to the seasons, already tells us that they will always over hedge during the good times and under hedge during the bad times, because that’s always what happens when you use a static hedging strategy.

There are two main mistakes that endowments and other big institutional investors make, which is to not invest according to conditions and to over diversify. This over diversification is seen in both the types of assets that they invest in as well as the particular diversification they seek within an asset class such as stocks.

It’s good to mix things up somewhat, but we need to only do so when it makes sense to, and in particular, makes sense to do now. Having most of your funds in bonds would be a bad idea right now, but there have been times where this has made a lot of sense and times where investing in either stocks or bonds would not produce a desirable risk-to-return ratio.

Harvard is happy enough with their 6.9% no doubt, and may actually be risk averse, but if this is the case, they need to make sure that they are protecting themselves more when more protection is needed, and choose less protection when less is needed.

The main worry here is that it will start raining and their current asset diversification which had them missing the mark by this 31% over the past year will leave them under protected. They will beat the market no doubt but that’s little solace when you still take a lot bigger beating than you should if you really were managing risk better.

We need to modernize what we call modern portfolio theory and actually make it more responsive to what’s actually going on, the present risk and return calculations, and make the theory more dynamic and responsive and therefore more optimal.

The starting point with the current theory has us entering in historical data of far too wide of a range, both in terms of what we are estimating expected return and risk to be. It seeks to optimize for these values, but since the values are longer-term averages, the data set is too old and too irrelevant, and this is where the theory fails.

The fact that we think that we can predict the return of future years very well looking at returns averaged over many years shows how out of touch this theory is, and it really doesn’t matter if we look at the average return over a period of decades, as that isn’t going to tell us very much and neither will our using variance numbers over such a long time period. Ironically, the model has way too much variance over the life of the data set to be useful and this is a major flaw but one that remains virtually unrecognized by mainstream investing.

The solution would be to use a similar formula but with a much shorter time frame as well as using this dynamically, one that would add hedging to the degree that the risk actually is increasing now, and have us more focused on returns when it is appropriate. If we use updated numbers like this, we can ensure that we are making decisions that are much more connected to the present, where more recent information is afforded a more appropriate weighting, as we should, which would be far more than it is presently given.

Fixing this will require a bigger revolution in investing than the original modern theory inspired, as it will require that we throw off the shackles of looking too far back upon the past, and these shackles are on pretty tight.

Andrew Liu


Andrew is passionate about anything related to finance, and provides readers with his keen insights into how the numbers add up and what they mean.