Jim Manzi replies to me on Apple cash including a post I previously missed. I am going to reply to Jim’s points not necessarily in order.

He writes

I have no idea whether Apple stock should be a buy, sell, or hold, but if Smith is right that the current shareholder base of Apple massively misunderstands the true capital requirements of the business, then he has a huge moneymaking opportunity. If he really believes in his investment thesis, he should borrow a lot of money and short Apple’s stock.

So, I am not actually making any statement about whether Apple’s stock price is “too high” or “too low.” I have a general thesis about the corporate commons but here I was actually responding to the assertion by many folks that Apple’s stock price is surprisingly low and has relatively speaking not kept up with the company’s growth.

My original case is that Apple has increasingly signaled that it intends to burn lots of shareholder value and one has to consider that when valuing the stock. Or, to put it another way one has to consider that the stream of discounted future profits includes a considerable number of negative values and Apple has signaled that it is not taking steps to minimize the number of negative values.

Since, then I posited the reactor model

His most recent post on this subject develops an analogy between the workforce of a tech company and the particles in a sub-critical fission reactor. This is meant to be literally (as far as I can tell) a sketch of a mathematical model for why Apple requires a huge amount of cash on hand to retain its employees. At best, it is pure speculation. And based on any practical experience in a tech company, it’s also extremely implausible that Apple would start to shed important engineers, or be at a disadvantage in recruiting, if it had built up, say, $40 billion on the balance sheet instead of $80 billion.

Its totally true that my model sketch is pure unadultered speculation. However, I think some folks are missing how this would work in practice.

Its not that if Apple had say $10 Billion in cash that some engineers would wake up tomorrow and say, “I don’t know guys I really think I would do better some place else”

Its that 18 months from now, when previously unknown 15 year-old engineer Jenna Markozy invents Gablet, a device that controls your entire professional and personal life by you making simple and highly intuitive clicking sounds that Apple could be in trouble.

If it has only $10 Billion then it will watch as all the top engineering grads head to Gablet and slowly its best engineers bleed over there as well.

If it has $70 billion then it either offers Jenna $30 billion for Gablet or it sets aside $60 Billion to build a Gablet destroyer. In either case it gets keep offering engineers the prospect of working on the Next Big Thing.

Yet, unless Apple got a deal on Gablet or you appeal to “synergies” then its shareholders do not benefit from the purchase of Gablet. Apple simply transferred the discounted present value of Gablet’s profits to Jenna and in return received ownership of the company. At face value this is a wash as far as shareholders are concerned.

Though there remains the strong possibility that the growing Gablet division comes to carry the infrastructure for the rest of Apple, which is equivalent to Apple slowly burning itself down.

Shareholders on would have been better off to simply have Apple dissolve itself and take the cash and invest it Gablet. Or, equivalently for Apple to buy Gablet and then dissolve the remaining parts of Apple. However, how likely do we think either of these scenarios are?

Now suppose that Apple spends its hoard building a Gablet destroyer. This opens the possibility for pure burning of the corporate commons. You might say well no this is just competition and that’s good. Ok.

But, do you imagine a conversation at Apple headquarters that goes like this. We have a decent chance of developing a Gablet beating product. However, if you look at the resources involved the net return is doesn’t make risk-adjusted sense. We should simply continue to produce Iphones and Ipads for the next year or so until Gablet completely drives our sales to zero, then dissolve the company and payout to shareholders?

This conversation seems absurd. Yet, there must exist some set of numbers that makes this the optimal choice. If no one ever chooses it then that is prima fascia evidence that they are not optimizing.

So, that’s how the model would workout in real life.

Jim also says

I’ve started and run a pretty successful enterprise software company, and I generally held a lot of cash on the balance sheet. From the perspective of shareholders, there can be many good reasons for this. First, do you know when cash-on-hand is most important? When nobody else has any. You can buy up the best talent, patents, and assets when they are cheap; you can make big technology investments when they are cheaper; you can make big marketing pushes for the resulting new products when competition for customer mindshare is lower, and so on. When times are good (or at least not catastrophic) it seems like you could always get your hands on cash when you needed it, but that’s least true when you most want it. Cash is the option to act decisively at the moment when this can create large advantages for the company

Now there could be reasons why this could make sense and I have tossed them around before. If this conversation keeps going I will bring them up. However, I want to show why at face value this doesn’t make sense.

In Jim’s scenario he has a lot of cash on his balance sheet and that allows him to make strategic purchases. But, by construction you know who doesn’t have that cash – Jim’s shareholders.

If Jim had paid out the cash to his shareholders then they would have it. And, if Jim could convince them that he really did have all of these great opportunities they could give it back to him. The fact that no one wants to give Jim cash should be taken as evidence that giving Jim cash is not a good idea.

You could say “well maybe there is just not enough cash available period.” This is a monetary policy question and is important but so involved that we should take it up in another post if people are interested.

However, here is the key thing: Having cash for opportunities like this is always a good idea for Jim and for the prospering of his company. But, both the market and the economy are indifferent to this. Jim is simply a vehicle through which resources are mutated into more desirable forms. Whether he or his company prospers is neither here nor there for the economy or for a properly diversified investor.

All that matters is the total return on capital and that’s the question we are trying to get at. There are decisions which undoubtedly help an individual company prosper but undoubtedly drive down the economy wide return on capital.

Apple is apparently holding the cash outside the U.S., so another possibility is that they’re playing for time before repatriating it because they think corporate tax rates might come down. They might be playing any one of a million tax angles.

Yeah, this is just akin to saying that I am reading their signals wrong which of course could be true.

Another possibility: A massive cash pile can discourage potential competitors from entering important markets, because they know you can retaliate by either crushing their foray into your territory or by going after their cash cows. The U.S. will hopefully never launch its nuclear weapons, but we use them every day.

Again, its clear why this is good for Apple. Its not immediately clear why its good for a diversified shareholder who also own stocks in the companies that Apple is deterring. Though, you could make the argument that this helps tacit collusion among large firms are therefore generates monopoly profits for everyone. That is an interesting angle, though many of my IO friends tell me that tacit collusion never works out in practice.

First, big tech companies often don’t pay dividends for a long time, until they do. Sometimes, these dividends are massive and continuing. Second, if there are continuing growth prospects for Apple that require cash (sometimes in ways that aren’t obvious, as per the first part of this post), then it makes sense for me as a shareholder to not want dividends for some time. The present value of the anticipated dividend stream is higher by getting more money later.

The links point to MSFT numbers. I want to do an add up – maybe a reader can do it for me – on what the future dividend and buyback stream would have to be to justify MSFT’s share price at various times discounting by the AAA corporate bond yield.

It may very well end up being that MSFT buyers at any point will have walked away to the good. We shall see.

What’s not clear, however, is that MSFT shareholders are better off having not gotten the money sooner. Mixing MSFT assets with cash reduces the options for me as a shareholder. I could want a bigger stake in MSFT real assets, in which case I would buy more stock with my dividend. I could want the cash somewhere else.

However, forcing a group assets – Microsoft real and Microsoft owned T-bills – to be bought and sold in a particular ratio makes it harder to achieve the portfolio balance I want. It may be the case that with a significantly large portfolio this constraint is non-binding but I would have to think about it.

What seems clear though is that the shareholder is unlikely to gain by imposing this constraint.

I don’t believe in anything approaching purely efficient markets. But any time a journalist or academic claims that some company could create enormous value by taking some simple action, the obvious question to them is, “Why aren’t you a billionaire?”

Key in my claims is that under none of these hypotheses is it in the interest of the Management to take these actions. Whether its in the interest of shareholders to take some action different than what they are taking I am not taking a stand on.

What is true – and a puzzle – is that under some variants of my claims it would be in the interest of private equity to take over the firm. The private equity problem is difficult as well though. Because, its clear how given these principle-agent problems one can create discounted present value using private equity.

Its not perfectly clear how one extracts that value.

Determining the ideal control strategy for economic production is extremely difficult.