The Chain of Supply
A little lesson in supply logistics and a coming world of hurt.
(As usual, the article’s too long for Gmail; so if it’s cut off ya can find the full thing on desogames.substack.com).
I decided to write this article after seeing yet another message of shortages that haven't yet developed, but will soon, and i realized i had an opportunity to teach subscribers a lesson in supply chains. As i currently have an opportunity to show you how a shortage develops, before it actually comes to pass - but is within a process that has already started and will lead to shortages within a supply cycle or two.
In this case, it has to do with Cheese and Milk.
For a while now (on twitter), i've been highlighting the chart of Oats going parabolic. And yes, by now i understand it's hard to keep track of all my warnings of things going parabolic; but things are moving ever faster and the problems will only increase from here. It's the main reason i've opted for long articles and trying to teach my methods to others as much as i can, because the time where even i can't keep track of everything about to implode is fast approaching.
Hang on, i thought this was about Milk and Cheese?
Well, that's why the title is "the chain of supply". One thing leads to another. In my book "The Definitions of Value", i talk about "Utility Value" as being the most basic way of determining the value something has. Utility value means the value of the function of the item within society.
Or put differently, Water has value because it is Water. It performs the functions of water: It hydrates, it helps clean, it is used for all the things water is used for. When we try to determine the price of water, that only concerns trade: the amount of goods/services/money exchanged to acquire water. The reason why we want Water in the first place, meaning the reason we're willing to spend money on it, is because of the functions it performs.
Dirty water can't clean, so it has less value then clean water. Less people will naturally desire it, thus its price will naturally end up lower. This precedes supply and demand: Because less people desire it, they won't ever come to demand it, nor desire to supply it. To simplify a long and complex book.
So, when you're looking at supply chains, it is extremely important to keep the function of the thing you're looking at in mind. Price is not the end all be all. Price is a derivative from supply/demand, perceived value, front-running and a whole bunch of other stuff. When it comes to physical supply which you need, price is the last concern. Obtaining the item is the very first. As China recently learned when it comes to coal. And while drafting this article the message came out that they've placed price caps on thermal coal. Meaning it'll run out not too long from now, as all price controls do is disable the "pricing out of the market" mechanic, aggravating shortages rather then fixing them, turning a slide into a crash.
Hence that i've said multiple times that the Oats chart going parabolic is the chart that has me worried the most out of any chart going parabolic right now. Copper's a problem, sure. Aluminium and Magnesium too, sure. But we can live without cars if needbe, electric or otherwise, and prioritize the resources left to where we need them. We cannot live without food. It's a necessary expense, and we'll pay whatever price we have to - or reduce the amount of mouths to be fed.
Oh sure, not that many people eat Oatmeal. The oat prices going up doesn't affect us that much. Which is what the lesson in theory above is for. If you're smart or curious, by now, you must've wondered "well, what is oats used for then?". You'll find virtually nobody in this day or age actually googles questions like that, because if you do this is what you find:
From this one search it's easy to surmise not a single goddamn policy maker uses any search machine to look up anything, or they'd be freaking the fuck out about the oats chart going parabolic just like i am.
Yes, it's not the only livestock feed. But having one source of food fail puts stress on other sources of food. If those sources can't adequately scale up to fill the void - chiefly because their prices have gone up too - there's only one option left: Less mouths to feed.
Which is the tweet i saw pass by not too long ago that prompted me to write this article.
Ah yes, now you see the connection.
Oats prices are running high, causing the feeding of livestock to become more expensive. Soybeans could be used to compensate, except those are at 7 year highs still. Corn is usually given to cows for slaughter rather then dairy cows, as Americans prefer how it marbles the meat, and beef prices have already risen so that's where that supply is going, increasing the herd meant for slaughter. So the only option left is to reduce the size of the rest of the herds remaining. Naturally, less dairy cows means less milk produced and less milk to make cheese from, so those prices must go up.
That it hasn't happened yet, hasn't made that statement already true if the current situation is allowed to run it's natural course (I.E. nobody intervenes).
Higher prices for Milk and Cheese would normally compensate for the higher feed costs for the cows, except that's not how this crisis is getting started. Normally it comes from the top - consumer demand increases, which reduces currently available end-product supply, which increases prices that then increases producer income to expand supply with and stabilize prices again.
This crisis is coming from the bottom - the input costs for livestock, the feeding cost, went up. This causes a contraction in supply at similar demand, which causes prices to rise. The higher prices will allow producers to afford feed better, but that's not the problem here - the problem is there's a shortage in the feed. Whether or not producers can afford it doesn't matter if there's nothing there to afford.
As a result, thanks to the supply cycle mechanic - meaning feed needs to run out before it impacts cow stock, which has to go down before milk and cheese prices rise - Milk and Cheese are still at historical averages right now (on the longest scale the charts offered me) and provide an investment opportunity.
Which is another good show that inflation is not transitory, and can't possibly be - infact it must get worse, because increased milk and cheese prices haven't been included in the CPI yet; they haven't gone up yet! But, if the number of Dairy cows in the US is dropping at historically fast rates, necessarily, Milk and Cheese must become more scarce within a few months of time as inventory is drained but new supply is disappearing instead of coming online - which'll continue until prices have risen to such a degree as to make new supply profitable ontop of the increased prices. And it will impact the CPI. This is not a question of If, but When - and even the When can be estimated pretty accurately at this point.
Just to take a small tangent now that i'm talking supply cycles, that other beef feedstock Corn just bottomed its supply cycle at 7 years highs, meaning the next spike starts from this price as a base - meaning the spike above the last one will be that much higher if more supply isn't planted in the ground. Normally that'd happen, but with Fertilizer prices also being squeezed due to a lack of natural gas, don't plan on it. And that's a huge problem nobody is expecting either, since corn is used in alot of human foodstuffs within the US, as well as livestock feed.
The problem should be clear from this one chart alone. If farmers haven't increased the supply of corn with the funds the price-spike afforded them, god help the US on the next one. At the latest, it'll go up towards the end of May.
Luckily, we've never had this many neat data sources available on the internet, and there's a chart for everything these days.
While the amount of corn planted is near historic highs, especially compared to the late 80's, it's not considerably higher then last year, or even 2012/14 when prices collapsed. (funnily enough, the total amount of acreage planted also ran up during the last period of heavy inflation, going from 67M in 1972 to 84.59M in 1976, and staying around that number until 1985).
From this it's enough to surmise that US corn supply won't increase by the amount implied by the price, atleast not before May, meaning it's highly likely the next price spike will be pretty epic. The acreage just doesn't exist.
In any case, i'm keeping this article uncharacteristically short (shortest i've made so far at ~5,000 words) because the problem should be obvious to anybody with even half a brain, and i don't have to write up a whole history of what happened if the event is still ongoing - y'all can follow it with me! I can still be wrong as futures might get capped along the way, atleast in the case of Corn. But if the actual supply is what's needed, that's just going to lead to more shortages. Price controls never work in times of crisis.
To add a bit of length to the article, and in the spirit of teaching, i'd like to recommend something to everybody. Especially to the next generation of political leaders who are still growing up.
Go play a videogame. One specifically: Factorio, or its derivatives such as Satisfactory or Dyson Sphere Program will do, depending on personal taste.
The reason i mention this is...well... that's how i learned supply chains and logistics. I've played those games for quite a bit of time, so stuff running out has become a secondary nature for me.
These games revolve completely around pure logistics and supply chains. The premise is very simple: You crash-land on a planet somewhere in space (a theme with all 3 games mentioned, a genre staple by now), and you have nothing.
Literally nothing. With no one else around either. The objective of the game is to create ever more complex components which add up to the ultimate objective of the game, whatever that is. You start by mining Iron ore and Copper ore - manually, because you have nothing - in order to get some initial resources. You can craft virtually everything yourself in the game, but creating resources in bulk is time-cost prohibitive. So the game is played by creating automated production lines.
Automated miners produce packets of ore, which inserters drop ontop of a conveyor belt and pick back up from it again to "insert" into other buildings. These buildings perform a function automatically as well - in the case of Iron ore, it needs to be smelted into Iron plates, so the ore needs to be inserted into a Smelter - which transmutes whatever resource is put into it to the next level. In the case of Iron Ore, this is Iron Plates mainly - though pure ore is sometimes also used for other products.
Once you have iron plates and copper plates, you can combine them. Don't worry - the games follow a very basic intuitive "resource tree", so you don't have to "refine" the ore or "filter out impurities" or anything like that (though if you wanna be a masochist, "realism" mods do exist for Factorio). Playing one of these games through to the end to see "what resource comes next" also gives you a decent enough understanding of what product uses what resource in real life. Good enough even that, if i was teaching logistics in university, i would make it a year-long project for students to create the biggest sustainable base they could with the expansion mods (without enemies) and optimize the logistics as best as possible for extra credit. For anybody else i'd recommend the base game though, as that's optimized to be fun, not brutal.
From this point onwards it becomes a numbers game. Iron ore is being produced at a rate of 30 a minute, smelters smelt 5 ore per minute, how many smelters per 1 miner to fully saturate production?
Naturally, you can just brute force this by overcreating the number of smelters you need. Don't think just because i'm a genius i spend my time doing math; i play games to have fun and then if i can pick up knowledge through pattern analysis, its a bonus. Still, art imitates life imitates art, and it'd be foolish to think a digital simulation of a specific part of life doesn't contain alot of knowledge on the subject.
1 copper and 1 iron miner gives 12 smelters, 6 each, to create 30 plates of either a minute. Except the item i need requires 2 iron plates, and 1 copper plate. So optimally, you'd make 2 miners with 12 iron smelters - but for that, you need the resources to be available in the world. Maybe you only have enough iron ore veins to place 1 miner. And these veins aren't unlimited. Running out of resources is a common thing in these games, as it's the mechanic that drives expansion out into the randomly generated world.
Now, this is just to give an overview of the games and their core mechanics. How they actually play is as follows:
There's not much of a problem while setting up your basic base. You'll get resources and power going, and you're working on expanding into a new production line when you run out of personal resources to craft buildings with (which are generally more efficient to craft then to assemble). So you go back to the rest of your base to collect some from chests that siphon off materials from belts for this purpose.
Except there's far less of them in the chests then you expected, and you can't expand further. How the hell did that happen?
So you go check these resources. 10 out of 10 times, the conveyor delivering one resource is either completely empty or only sparsely delivering resources, while obviously you're trying to keep these fed 100% of the time to max out production. You'll need a LOT of resources in these games.
So you go check the production line of that resource. But you'll find - it's not actually that resource which is running low. The production lines going into that resource are already stressed and aren't delivering the maximum amount possible, leading to the intermediary product being produced less then wanted, and that is causing the finished product you were looking for to stall in production.
And that is the key lesson to take away from games like Factorio, the reason why i don't have problems spotting these supply links and crunches. Because these games are based around filling these conveyors up to the max with resources, and then working off that, with barely any overcapacity left until you yourself expand production, which causes problems because nobody can keep everything in mind all the time. Which closely mimics our real life just-in-time economy, except there the workload is divided between many humans.
And then you run out of iron ore, which gets consumed in ridiculous amounts. Or power. Or coal to feed the power plants. And any of those shuts your base down completely. Especially when it's power; getting your base started back up if you're using base power to power your coal miners which deliver the coal to power your base is a bitch. If you lose one of the most basic resources, you lose every other resource based off that basic resource as well. Which is why i've got chests siphoning off one of each item everywhere so i always have the resources to restore the base - just in case inventory.
Or you don't even have to run out: You had a belt of iron plates which was continually filled, so you added a splitter to feed those plates into a different section of your base - but because the resource you're building there also uses another resource which by itself also uses iron plates - suddenly there's a heavier draw on iron plates from 2 sources rather then 1, and the belt stops being saturated. That then suddenly starts impacting everything else being fed from that belt, and parts of your base you thought would run fine shut down.
You'll spend alot of time in these games fixing your own mistakes and balancing out production across sections of your base while building up to the next resource you need to produce. And by doing so, you'll get an innate feel for supply chains. Whenever you see a supply problem, the question of "oh dear god what part of my base blew up now?" becomes a reflex.
Uhm, the question of "What else might be affected". Don't forget to take breaks while playing!
To show this, as well as the real life parallels that can be learned from videogames, let me take you to a random supply problem in the base in one of my games.
What you see here is a screenshot of Dyson Sphere Program (and yes, the bottom left shows a minimap of one half of a planet covered with production facilities). In this case, the orange buildings are Mk.1 Assemblers, set to produce an item called a Particle Container. This item requires 3 other items to be constructed; the details i won't bore you with.
These 3 items arrive by conveyor infront of the building, while the finished product comes out of the back (the green arrows are Inserters). As you can see, 2 belts are completely saturated with resources, meaning i've got overcapacity there at these production levels. It's the 3rd which is the problem, and the greenish item shown on the selected section of belt is called a Magnetic Motor. Clearly i need alot more of them because my production is being held back (or i used to have more and ran into trouble).
Finding supply problems in these games is pretty easy, just follow the conveyor back to where-ever the hell you drew this production from. Oh sure, you'll end up remembering maps in some detail (it's a good memory trainer too), but never entirely. Chasing belts is a common thing.
Here's where that belt leads too, my magnetic motors production facility (for this part of the planet cause there's a dedicated one on the other side too). As shown, it's not actually the magnetic motors that are constricted in production - it's a subcomponent called Rotors that isn't being produced enough. The production of magnetic motors doesn't need to be increased, instead increasing the production of rotors should do it.
That's another good lesson for real life; if you have supply problems in one section of the economy, introducing extra intermediary capacity isn't going to work if the problem is with subcomponents or raw resources. It'll only aggravate the problem by creating more sources of demand. Which is why from the start i never believed the transitory inflation narrative, because it's impossible that all parts of the supply chain scale in unison. That's simply not how supply chains work.
As stated, it takes time before the reduced dairy herd reduces cheese and milk supply far enough to increase prices to stimulate increased investment in dairy cows. But if the input costs for cows, the feed, is going up at the same time - you need milk/cheese prices to exponentially rise in order to increase the supply. A small overview shows this:
Oats: $10, Cows: 10, Milk/Cheese: $5, Demand = 10 cheese.
Oats: $20, Cows: 10, Milk/Cheese: $5, Demand = 10 cheese.
Oats: $20, Cows: 5, Milk Cheese: $5, Demand = 10 cheese.
Oats: $20, Cows: 5, Milk/Cheese: $10, Demand = 10 cheese.
Oats: $20, Cows: 10, Milk/Cheese: $10, Demand = 10 cheese.
If you then create extra demand, by say printing alot of money and force feeding liquidity into the economy while goods start running out, Milk/Cheese need to go to $15 to get 15 cows at a price of $20 for oats and a demand of 15 cheese for supply to stabilize with the increase in demand from money. While we started at $5. If at any point, demand drops, supply stagnates.
Obviously the numbers are arbitrary, but that's merely to simplify the supply chain for the example. The real world is the same loop repeated with alot more components.
To go back to our digital simulation as it cuts out the element of time to a large degree (supply cycles are effectively instant ingame). We know we need more Rotors, so we chase a belt again.
I've followed the belts to its source, and as you can see, here we've run into a production constraint. Since the belts are completely fed, it's not a problem with the sub components. Instead, i'm just demanding more rotors then i can supply, so lack of production's slowed down the line.
I need to expand this production, but i've kinda run out of space in this part of my base. Luckily, since these games are based around massive production and i'm still using Mk.1 assemblers/inserters, i can simply upgrade everything to Mk.2 assemblers, which produce 50% faster each.
Which immediately shows why these games make you think deeper about supply issues. And mind you, i didn't specifically build this situation for this article - i literally just fired up my normal game and went looking for a problem cause i knew there'd be one.
Here you can see Mk.2 assemblers with Mk.3 inserters. Production's increased somewhat, but it's not nearly optimal. Suddenly, the last belt is empty. Because several assemblers before these last ones in the line were upgraded in production speed, they also started eating more resources. And apparently, the overcapacity on the last belt was just enough to feed the available production, so i've gained not nearly as much as i thought. The assemblers lost compensate mostly for extra production the others gained.
This is just about the best possible explanation i can give for what the hell happened to global supply chains over the past 2 years. Why it seems that everything is going up in unison.
Games like Dyson Sphere Program or Factorio cut out money. There's no one to trade with, it's about creating supply chains. If there was a way to use money to import resources (for Factorio there's probably a mod), the price of the item on the empty belt there just spiked through the roof. There's overcapacity in manufacturing demand, but not nearly enough manufacturing supply capacity to produce the item.
We can relate this immediately to the situation with Corn. The average price of Corn per bushel (per packet) might've leveled off at ~90% more expensive then the average over the last decade, capacity can't suddenly increase by 90% to compensate. Which is required to bring prices back down at the same supply of money. Since there's still plenty of liquidity, and acreage planted has only gone up by ~3%, prices must continue to go up.
This is what is meant by "pricing people out of the market". As long as that conveyor belt doesn't force the last 3 assemblers out of business, prices will rise, because there's 10 assemblers demanding supplies only available for 7 of them.
Except that the game doesn't use funds, so it's much easier to break the stalemate and learn a valuable lesson (and to hide from finance for a while). In this case, production's fine, it's simply the conveyor belt speed that's lacking. The 1st belt is already Mk.3, but the last 2 belts are still Mk.2, including the underfed one.
This is easily fixed, and in this next screenshot, i've done nothing else but upgrade belt 2 and 3, and wait a while.
Nothing's changed. I'm still producing about the same amount, the back 3 assemblers are still running out, only now - my 1st belt is empty, which is already Mk.3.
Or in short - i can't upgrade belt speed further. Mk.3 is the fastest one. I need more basic resources. And i don't have those in this part of the base. I have no choice but to expand facilities within a space i don't have.
In the game, this isn't a problem as i can just drop another orbital transport station there and import them from the nearby planet. But it's clear what the real world parallels are here: If this was real life, and this planet was all i had - i'd be screwed. Production cannot increase, and i must find additional sources of Iron to create more Iron Plates which this line now requires, which don't exist.
And i'll tell you one better: The item on the first belt is not used at all in the production of the item on the 3rd belt. They are two completely different production lines. So, what happened here?
I increased resources on the 3rd belt is what happened. I didn't do anything different from what i've described. But, since the 3rd belt became saturated again and allowed production to increase, the second belt also started moving again. And the item you see on the 2nd belt are Iron Gears - naturally made from Iron Plates, which are also on the 1st belt.
Because i saturated the 3rd belt, more iron gears could be consumed off the 2nd - which increased the resource demand on a production section before this section, as more gears needed to be made. This started draining more iron plates by itself, which left less plates available for the rotors. And because the gears are being made before the rotors, the gears belt is saturated and the iron plates - the input resource for gears - is not.
Ironically, the solution here (aside from adding plates), would be reducing gear production by a little to balance things out. Even so, i doubt i can feed the last 2 assemblers here without scaling even more production up, without those resources being available.
In this situation it's clear: While i thought the production constraint was in Magnetic motors, it turned out to be in Rotors, which itself turned out to be limited by the supply chains running to the very beginning in my base, to raw resource production. I had not nearly the overcapacity i thought i had. And neither does the global just-in-time economy. Oats are the iron plates to the Cheese's magnetic motors.
In this case, there is no answer except off-world expansion. The game was designed for this of course, but playing these kinds of games has left me with an existential crisis of sorts of "where are we going to get the resources for infinite expansion?". Economies contracting is considered Verboten and any sort of recession these days is frowned upon, with even 5% corrections in markets getting the Fed's liquidity hose to be put to a halt.
But infinite growth is impossible, and life isn't a game. In the case of Oats and Corn, there isn't some sort of technology we can research to immediately boost production by 25%. Only more acreage will do that, and there's a limited amount of that available. If then all prices go up at the same time because the item we use to exchange for food - money - becomes more numerous, we cannot expand production that far that fast. Prices will and must go up, and not over the span of 1 year either. More like a decade, if it's possible at all to raise production that far, and the economy doesn't collapse into hyperinflation under the weight of all that liquidity (for that i refer to my previous article).
Whenever you expand one sector of the economy, a bottleneck develops somewhere else - and another sector of the economy withers altogether as it has resources pulled to develop the one sector. Once we run into the limits of production, and the price charts say we already have - there's no option left but reducing demand through pricing people out of the market.
And in the case of food, this means, people starve. Can't afford food, you're not eating food, you die.
And yes, running low on power because you plopped down a whole bunch of new manufacturing facilities is a frequent occurrence as well in these games. Yknow how you solve those?
Nuclear power in Factorio/Satisfactory, Fusion power in Dyson Sphere program.
Because yknow, games have to actually be balanced to be fun, and solar/wind power never delivers anywhere near the power you need. Wind's entry level and Solar's decent mid-game (especially Factorio where pollution attracts very angry aliens) but doesn't work at night, so all of these games inevitably end up with some form of Nuclear power as the only thing that makes sense to deliver the massive amount of power your base needs for advanced materials.
But yknow, life isn't a game, so the world replaced cheap reliable coal with intermittent "renewable" power while shunning nuclear and decided to make up the difference with gas without actually planning for or investing in additional sources of gas, and then printed a bunch of money so everybody could afford power-intensive products ontop of it.
I really wish people would play games a bit more. Not that everybody needs to become a logistics expert. Just enough to realize what happens when you increase production in one part of your economy, but not the connected parts.
Ahead of time that is, and not at the register. Maybe then people will stop voting for idiots who haven't played these games (or used a search engine) and who keep demonizing supply while stimulating demand, while being rich enough to afford any price themselves. I still can't believe people expected this to work. What did y'all think was gonna happen?
- Kirian "Deso" van Hest.
Find me on Twitter! https://twitter.com/DesoGames
Twitch: https://www.twitch.tv/desogames
My own website: https://www.desogames.com/
And buy my Ebooks for alot more content like this!
The Definition of Money: http://books2read.com/u/bzdaVz
The Definitions of Value: http://books2read.com/u/3yaZWv
Ethereal Value and the Cryptofuture: http://books2read.com/u/bMwrNA