Cash Is the Operating System

The Maxims #

I remember when I came to the conclusion that the main thing about finance, despite all the other ways you can look at it, is that it's primarily all inflows and outflows. Or opposites and them balancing (or not balancing). Credits and debits. It's a useful way to distill something terribly complex, with infinite moving parts.

One week I had approved two vendor contracts, both strategic, and if anyone asked, both defensible in isolation and together. In my personal life, I don't like to spend before I've earned. But in business, a lot of times you spend to earn. Each approval wouldn't even come close to breaking the bank, but it’s a strange and specific sensation to constantly mull a multiverse of possibilities: how much room is left, what if xyz customer delayed their large payment, or what if some unknown but obviously more important use of the money suddenly appeared.

Grow revenue faster than expense, spend less than you take in, keep enough cash around. FinTwit infinitely shoots these maxims at us, and all of them are largely true in the same way "eat vegetables and exercise" is largely true, until we visit the doctor and learn we can't really absorb Potassium or whatever. But we have to decide what to eat right now and how to fit in that run. Do we hire someone, renew something, replace the roof, go on that trip, pay off the loan, hoard cash because all of a sudden the world has gone insane, whatever.

Kernel Logic #

I once heard something like "finance is a medieval priesthood, but everyone can learn to speak Latin." Which obviously I love that analogy. You can see all the church-like ceremony, the paternalistic pressure to respect it at a distance and trust those who can understand the revealed truth better than you.

When I heard that sentence it was like an axe to the head. It's not one of those fields that is so indiscernible your head hurts, like theoretical physics, though both fields sometimes have to deal with quantum states within a black box.

Personally, I prefer a computer metaphor and just think of the idea of finance similar to the dreaded kernel, which occupies very little of your brainspace on a given day but is absolutely essential to everything going on above it. Most of the time, the system hums along in due course: invoices clear, payroll goes out, and vendors get paid on time or close (I once worked with a consultant who wanted all Accounts Payable to be about 40% overdue). Debits and credits balance out. When it's working invisibly, you can feel that it's healthy.

Until the spinning beach ball or the blue screen of death and suddenly it doesn't work invisibly. And one morning a message appears:

Hiring is paused pending review.

That message or anything like it ("Everything must now have a Purchase Order. Please reach out to so-and-so...") is like when your computer sends you an interrupt. Uh oh. And the speed with which that interrupt ripples through an organization is something that still genuinely startles me, even though I've been on both sides of that swing multiple times and should know better.^[1]

The kernel metaphor is useful to me despite the obvious over-reaching going on. I'm skeptical by default when people reach for engineering analogies outside of real engineering stuff, because usually when I do this or I see anyone do this, it seems like we're flattering ourselves into thinking we're doing something deeply precise and systematic. Real kernels follow explicit rules, which sounds obvious until you speak priestly engineering Latin and understand just how all those rules can interact. Humans aren't deterministic when deciding to part with cash, and so neither are organizations or any other groups of humans (see: US Natl Debt). A kernel doesn't approve a vendor contract because the relationship feels important or because saying no might damage something intangible that won't show up on a balance sheet. Again, I have to reach for a feeling because it's the best description I can find: it's like someone with insomnia once again checking the thermostat at 2 a.m., except the thermostat controls a lot of important stuff besides the temperature. And you can adjust it, but not quickly without one of those interrupts. So you're mostly watching it in the dark, at 2 a.m.

So why wander around at night and watch? Well, because cash gives you permission. Permission to do something, go somewhere, wait, change your mind, survive being wrong for a while, shift gears. This is basically the Wu-Tang thing: Cash Rules Everything Around Me.

Memory Leaks #

Time to stretch the kernel metaphor even harder: it's really hard to admit how often we create memory leaks ourselves.

Memory leaks are another useful metaphor, better than I originally thought, because nobody sits down and decides to have a memory leak. The program on your computer continues to do what it was asked to do. Each process has a reason to hold onto a little more memory.

Cash can go the same way because everything has a reason in isolation. The consultant, the hire, the renewal, the experiment, and the team asking for a few more months definitely have a reason. You can have conversations about them and agree to almost all of them. Until you add it all up, stop seeing it all in isolation and realize the total is more memory than you want to use.

And I think it's because we don't really think that much about totals. We think about pieces, variables, the stuff that gets shoved into the equation.

I've sat in planning conversations where I agree with almost everyone. This or that hire, this or that new extra time or new tool. Makes sense. Then someone adds it up and the totality is off.

That's probably, for me, the strangest part of doing that kind of work. You can sit for an hour, agree with mostly everything, and still come down on the side of: "Hmmm, no. Not right now." ^[2]

It feels less like precision engineering than being handed the check at dinner, trying to equally split the bill but realizing three people ordered lobster.

Panic and Preservation #

Memory leaks are slow. The discovery of one is basically immediate.

Language is the first thing that changes. Where a week ago everyone was arguing about where to use the cash, what to fund, who to hire, where to go faster, now everything is under a higher level of almost Socratic-level scrutiny. You discover and rediscover the word "essential," again and again.

It can feel surreal because it makes sense and it's immediately different than how things were before. It's good to be slower now, to have more eyes on exactly where cash is moving. To argue about "essential" and "necessary."

And then the controls start multiplying, breeding. More signoffs and thresholds. Soon enough, the issue can be, often is, this: does anyone remember what all the saved time and money is even for?

I understand why people roll their eyes at the word "controls." I also think they're mostly wrong to. But I get it.

As most people encounter them, controls feel like the accumulated bureaucratic sediment of every bad thing that ever happened. Or a policy that now applies to everyone regardless of whether they were involved in the original bad thing. Controls can feel like the organizational equivalent of those signs in hotel rooms asking you to please not do something insane, like iron your clothes directly on the bed. Someone did this. Now there's a sign.

But every control that exists, even the ones that seem absurd, was born because something somewhere went a little or a lot wrong. Someone put their hand on the organizational equivalent of a hot stove, and now everyone's wearing oven mitts. And the reason finance people defend them, even when they personally find the controls cumbersome, is that they've seen what happens in their absence.

Controls are binary in a way that judgment is not. A control either exists or it doesn't. It applies to xyz situation or it doesn't. There is no expense approval threshold that says "use your best judgment based on the current cash position and the strategic importance of this particular spend and compare that (quickly) to all the other demands on the same pool of money and see if could be used better elsewhere." Good luck figuring out how to make that a threshold a human can understand. So instead you set a threshold somewhere, forget it for a while, and accept that it will be either too high or too low depending on the situation, but always annoying.

Few people want to be the person encouraging more danger, to say, "Hey, I think we're all clear and let's get a bit more dangerous again." It feels weird to remove one when they're just easy to keep around. And something bad happened in the first place because it wasn't there, and it feels obvious and responsible to keep it because what if that same bad thing happens again? You look around, don't see the bad thing, but you inherently know the risk.

So a lot of the controls just...stay.

Eventually cash is protected so much that the uncomfortable question is what exactly is it for?