This is an excellent chart. (Archive here.)

The diamonds in the graph below show the amount spent each year from endowment, while the bars show the endowment’s value at the beginning of each fiscal year. Over the past decade the amount spent from endowment doubled from roughly $40 million to $80 million. The colored bars indicate the three fiscal years, including the current one, that began with the endowment at $1.4 billion. In the first of these years (2000-01) we spent from endowment almost $40 million less than we plan to this year. In 2005-06 it was almost $20 million less. Some of those differences have been made up by increases in revenue from student fees and from gifts, but the rest is coming from a level of endowment spending that’s not sustainable over time.

1) I have made the exact same point again and again. Morty’s Armstrong kegger spending spree has Williams using twice as much ($80 million versus $40 million) from the endowment this year as it did a decade ago. Reasonable people might not blame Morty for this since we all (?) thought that the bubble would go on forever. Yet there is no excuse for not adjusting to this reality as soon as possible. Security is here to end the party.

a) Jeff Zeeman comments:

We’ve had this debate a zillion times and I still say the same thing: Williams’ endowment grows from DONATIONS in addition to the market, as do all charities. … So 3 percent is ridiculous, because it excludes the former.

The way that virtually every single major endowment/school/foundation thinks about this topic is the same, even if you don’t understand it. Every trustee feels that she has an obligation that, if the endowment is size X today, she needs to ensure that it will be at least size X in real terms 10 or 50 years from now. How to do that? First, you have to ensure that investment returns at least match inflation. Once you have done that, you can take any real return (this is the mythical 5%) and spend it. You do not get to count future gifts in this calculation. First, they may not show up. Liberal arts colleges, even formerly strong ones, go bankrupt all the time. Second, if they do show up, they will be needed to replace the institution’s capital stock. That $25 million gift for a new building does not really add anything to the endowment because you have to replace the old building at some point anyway.

Now, you might be happy to assume that, because the College has raised hundreds of millions of dollars in the past, it will always be able to raise hundreds of millions in the future. Perhaps! And US home prices never go down nationwide year-over-year. And the stock market has almost always beaten bonds on a 20 year horizon. And . . .

Responsible trustees to not count on estimates of future fund-raising in making spending decisions today. They have a responsibility to not spend down the endowment.

b) Derek Catsam suggests:

No one has yet responded to my assertion that Williams will have hundreds of millions of dollars in endowments no matter what the college does, that those hundreds of millions will be more hundreds of millions in a few years, and that the best thing Williams could do would be to give raises, continue the no loans policy, and to use the opportunity to hire, hire, hire away.

This is the logic of I-really-want-that-big-TV-so-I-will-cash-out-some-money-from-my-IRA. After all, you have thousands of dollars in your IRA now! After you buy the big TV, you will still have thousands of dollars in your IRA. The big TV is excellent (we all agree). So, why not buy it?

The answer, obviously, is that smart people care, not just about the wonderful students/faculty at Williams today, but about the state of Williams 50 years from now. We want Williams to be the #1 liberal arts college then. How to ensure that? The single most important metric in 50 years will be the same as the single most important metric today: per student endowment. If you have a high per student endowment in 50 years, you will be able to create/sustain all the things (faculty/students/facilities) that make for a great college. If you don’t, you won’t. If Williams spends too much money now, it will no longer be the #1 school or in the top 5 or even, with enough recklessness, in the top 10. Fifty years ago (I think) Wesleyan and Williams were of equivalent quality. (That’s why they were grouped together with Amherst as the Little Three.) Now, not so much. Why? Because we got rich and Wesleyan got poor.

For my sins, I will continue to explain these obvious facts for as long as it takes. I disagree with the Williams trustees on many, many, many topics. But on the necessity of getting our spending from the endowment under control as soon as possible, they are absolutely right.

2) Note that this chart is highly misleading because it does not account from Williams’ debt. There is a difference, as Morty explained, between having a million dollars in the bank and no mortgage on your house and having the same million dollars but with a $500,000 mortgage. In the first scenario, you are twice as rich as in the second! But if you just look at the endowment, as this chart does, you miss the fact that Williams has (from memory) $80 million in debt in 2001 but $260 million today. What is a tripling of debt between friends?! (Previous discussions here and here.) The chart should be labeled “endowment” rather than “total financial assets” since the latter suggests that debt has been accounted for when, in fact, it has not.

3) Where do the numbers on this chart come from? I believe the historical endowment levels. But what rate of growth in the endowment is assumed going forward? That is the key number. (It would have been helpful to give the forecast bars in the chart a different color.) Also, the “assets use” number does not correspond to any of the endowment spending numbers that I have seen. Morty claimed that avail spending (dollars spent from the endowment on current operations) was around $90 million in 2008-2009. This chart shows it at just over $80 million. What is up with that? Presumably “assets use” has some special definition. Again, why can’t Williams provide a simple spreadsheet with the exact numbers of their definitions?

Print  •  Email