E-piphany

It is a lament I have heard recently from the manufacturing/supply side, but also an excuse I’ve heard from some of the economist ranks for why “this supply chain issue will all get sorted out; people are just going crazy.” In this column I want to explain why “overordering” is not only perfectly rational but actually demanded by some typical operational procedures.

The complaint is “our customers are not only ordering what they used to order, but they’re ordering far more than they used to. Basically, they’re hoarding and we have had to ration our product and only partially fill orders/only fill them for our top customers.” Now, hoarding is a real thing, but moreso for consumers than in B2B. There are, though, some serious reasons (by which I mean, ‘reasons that are held by serious people’) why it makes sense to increase orders at a time like this. And it’s not because you are assuming you’ll get 50% fill rates on your orders, so you order double in the hopes that you’ll get the actual amount you want. That’s an “unserious” reason.

One of the reasons I have written about before. Back in January, I wrote an article called “The Optionality of Inventories,” in which I predicted the companies would move away from lean inventory models because inventory serves as an option against bad things happening: if bad things don’t happen, you’ve paid a little more for your inventory; if bad things happen, you have a large gain (loss averted) because you had a cushion. That article is worth a quick read. I also point out that, as inflation increases, there is a financial incentive to hold larger inventories because the inventories themselves are increasing in value. To the extent that more firms are recognizing the option value of inventory, it makes total sense that the demand gets fiercer the closer to raw materials you get. The entire supply chain needs to hold more inventories.

But there’s another “serious” reason that is related to the length of the supply chain itself. “Fred, last year you only ordered 1,000 units. This year you ordered 2,000 units! I know your business hasn’t doubled. Why are you doing that?” Fred might well be doing this because lead times are increasing, and that mathematically increases his reorder point and quantity.

Reorder quantity mathematics, at the simplest level, is just “number of days of lead time” times “average use per day,” and you reorder when inventory declines below that number plus some “safety stock” which is essentially a fudge factor. So, if we are using one ton of flour per day, and it takes us a week to get flour, then we need to reorder whenever we get down to seven tons of flour (call it 8, just in case. That extra one is the ‘safety stock.’) And, when we reorder, we reorder at least seven tons of flour since by the time that order arrives, we’ll be down to one ton of flour. But if the lead time now stretches to two weeks, we are suddenly ordering 14 tons of flour even if our usage didn’t change.

That simple model works for very regular inventory usage patterns, but in many applications the quantity used (or demanded, if we are talking about holding finished goods inventories) is variable. In that case, the reorder point and quantity also depends on the variance in the order flow. Again, inventories are like options, and so the sophisticated way to think about the safety stock is the option value where the stock-out (you run out of inventory) is the strike price. If you want to never lose on that bet, you have to have a high option price (safety stock); moreover, the higher is volatility, the higher is the level of inventory required to maintain a given ‘acceptable’ level of stock outs.[1]

How does 2021 compare to 2019, the last time manufacturers faced “normal” order patterns that they are now seeing customers exceed? Well, there have been substantial increase in lead times, and substantial increases in every kind of volatility you can imagine. Lead times across the globe have increased probably 30-50% at least, and that means that required inventory needs to increase 30-50% at a minimum, plus more because volatility has increased.

So that customer who is ordering a lot more right now than they historically have is not doing it to “hoard.” They’re probably doing it just to manage inventory properly. Of course, that puts more pressure on the supply chain, and increases lead times further. It represents a one-time increase in GDP, as intended inventory accumulation adds to output in the period it is accumulated, and that pressure also boosts price pressure. And ‘round and ‘round we go.

And all of this, we should take pains to remember, started when governments decided to use cardiac paddles to resuscitate a patient they’d actively tried to kill, and central banks made sure they were hooked up to a strong current to do so. The fact that the body economic is convulsing should not be a surprise to anyone. The question is whether we can sue for malpractice.

[1] The only way to guarantee that you’ll never run out of inventory, if there’s any variance in the demand for the inventory, is to hold massive amounts of inventory. So in practice you have to pick an acceptable stock-out frequency, which enters into the calculation.