UR’s crash course in sound economics

Today we’re going to achieve a basic understanding of economics in one UR post. No previous knowledge of the subject is assumed.

UR can deliver this remarkable savings in both time and tuition because, and only because, our “economics” is an entirely different product from the standard academic sausage. We’ve considered changing the word—but this feels hokey. It also conceals our feeling that (a) for most customers, UR’s “economics” is a more than satisfying replacement for 20th-century industrial numerology; and (b) our version is far closer to the original artisanal craft.

Here at UR, “economics” is not the study of how real economies work. It is the study of how economies should work—in other words, of how sound economies work. Sound economies, as we’ll see, are also stable economies. All the King’s mathematicians have had some trouble in reproducing this property.

Since there are no economies on the planet which are even remotely sound, nor is there any prospect of any such thing appearing, this discipline cannot conceivably be empirical, quantitative, or worst of all predictive. Its only tools are logic and sanity. Sound “economics” can only be natural philosophy, not “science.” (A brand in any case increasingly overexposed—not least by its association with the King’s mathematicians.)

Moreover, since healthy economies are healthy by definition, there is no practical reason to study them. No remedy is required for their ills, since they are not ill. Thus UR’s “economics” is not only nonscientific, but nontherapeutic.

It does not even offer a path for transforming a pathological economy into a sound one. It prefers the old Irish directions: “don’t start from here.” If there is such a path, it is certainly neither gradual nor gentle. Since real governments are about as capable of following it as your grandmother of climbing K2, its existence is just as fantastic as our sound economy, and it is not even worth mentioning.

So why bother? Well, two reasons. One, sound economies are simple. Or at least, they operate on simple principles, which we can explain in one post.

Two, if you want to try to start to begin to consider understanding the giant bag of cancer, graft and rust that is a 20th-century economy, you may find it helpful to contemplate the imaginary sound economy inside it. Or not, of course. But at the end of the post, we will look at a few of the pathological ways to corrupt a sound economy. You may find them faintly familiar.

On to sound economics. Readers familiar with Austrian economics will find much to skim, especially at the start, but should also watch out for nontrivial differences in the origin of money and the structure of the loan market.

First: what is an economy? An economy is a network of conscious actors, who desire goods they can produce and exchange. All real human societies are thus economic. So are alien societies, if any exist, and if the aliens can think and plan. Virtual worlds which meet this definition are also economic, and display the expected patterns.

The desires of a conscious actor are unknowable by definition. We observe only the action. Moreover, the assumption of consciousness (which is rational by definition) tells us extremely little about these desires, especially when we start to order the desirability of goods. Which is more desirable: a one-carat diamond, or a six-pack of Fiji Water? Depends whether or not you are lost in the desert.

But rationality does give us one rule: that a good always has positive (if negligible) desirability. So it is always rational to prefer more goods to fewer goods. E.g., more money to less money.

This might be objected to in the case of your brother-in-law’s rusted-out 1981 conversion van. But the problem is one of semantics: his need to junk the van (which, not unlike our real economy, only needs a transmission, a couple axles and a little Bondo) is a liability or obligation, not a good. If he had a voucher for free vehicle disposal, his ownership would again be a pure good—however negligible.

We can state the principle more precisely by observing that a conscious actor always prefers more powers and fewer duties. A power cannot have negative value, because one can always decline to use it; a duty cannot have positive value, because one can always do it anyway. The classic form of economic power is ownership of some good; the classic form of duty is a debt.

And this is about it for ol’ “homo economicus.” What rationality does not and cannot tell us is anything else—specifically, what we really want to know, which is whether any actor X should prefer good A to good B.

Again, only X can decide whether to prefer the diamond or the water. Moreover, if all actors ordered all goods identically by desirability, we would have no economy by definition—for we would have no exchanges.

And it gets worse. We cannot even know all of X’s preferences—only those which are revealed. The only way for an objective observer to observe that X prefers A to B is if the observer actually sees X give A (or A, plus choices or minus duties) to some Y, in exchange for B (or B, minus choices or plus duties).

Does all this seem obvious? You’d think it would be. However, a considerable thrust of medieval pre-economic thought was the effort to determine just prices for goods, derived from objective factors. The urge to just-price theory is by no means extinct. We appeal to it, however vaguely, whenever we mention the “value” of a good, as if “value” were an objective quality such as mass or energy.

(Indeed the entire apparatus of price-index terminology, from “inflation” to “real” and “nominal” figures, “1980 dollars,” and the like, is deeply if subtly rooted in the fallacy of objective value. All this doxology is shunned by the sound. Indeed, the wise man would rather drop the N-bomb than let “nominal interest rates” slip past his lips.)

It would be an overstatement to say that all fallacies in economics proceed from the error of attributing objective value. But it might not be much of one. Objective value is the luminiferous ether of economics. There is no such thing as value (objective desirability)—there is only price (exchange rate on a clearing market). For example: consider your house.

But what are these markets? Perhaps it’s time to define them.

First we need to simplify our sound economy beyond physical reality, and assume that all goods are both perfectly fungible (each unit is identical, like a barrel of oil or a bushel of wheat—not like a house) and perfectly divisible (it makes sense to speak of 2/3 of a barrel of oil, but not of 2/3 of a Ford Explorer). Goods exhibiting these qualities are sometimes described as commodities—a confusing term, not consistently applied. In any real reality, most goods are not fungible commodities. However, once we understand the commodity-only problem, it is trivial to extend its principles to the ordinary sound economy.

For every pair of goods—say, oil or wheat—there is a market. This is some trading mechanism by which actors exchange oil for wheat and wheat for oil. For example, A may give B a barrel of oil, in exchange for six bushels of wheat. For this transaction, the oil/wheat exchange rate is one barrel to six bushels.

In revealed-preference terms, what this exchange tells us is that A preferred six bushels of wheat to a barrel of oil, and B preferred a barrel of oil to six bushels of wheat. Otherwise, they would not have made the trade.

But the interesting question is: why six? Why this exchange rate, not another exchange rate? For instance, we know that A would prefer seven bushels of wheat to a barrel of oil (because he must prefer seven bushels to six bushels, and he prefers six bushels to the barrel). We do not know if B would prefer a barrel of oil to seven bushels of wheat—but we don’t know he doesn’t. He certainly might. If he does, the same A and B could conduct the exchange at seven barrels to the bushel, and both walk away satisfied. Instead, they trade at six. Why?

The answer is that A and B get the oil-wheat exchange rate by participating in a market. Ideally, this market contains all players who wish to make this trade. The market clears when no trades are desirable to both sides—at the market exchange rate, no A and B can be found who want to exchange oil for wheat.

Of course, preferences and players are constantly changing, so the market never actually stops trading, and the exchange rate never stops moving. However, we can imagine the first day on which the market opens, with oil moguls on the left side of the floor and wheat barons on the right, all looking to establish an exchange rate.

This is the famous model of supply and demand. I dislike this terminology, because it implies some qualitative distinction between A and B—between trading wheat for oil, and oil for wheat. Of course it is arbitrary who is A and who is B, who gets to be the numerator and who has to be the denominator. But the terms are unavoidable. For this example, therefore, we’ll speak of oil-wheat supply and wheat-oil demand.

The oil-wheat supply is the number of barrels of oil that will be exchanged for wheat at a given oil-wheat exchange rate. So, for instance, the oil moguls might be willing to provide only 15 million barrels of oil for five bushels a barrel—but for seven, you will pry 40 million out of their hands. This implies that 15 million barrels are held by moguls who would take five bushels or less, and 25 million are held by moguls who will take seven or less, but no less than five.

The wheat-oil demand is the number of bushels of wheat that will be exchanged for oil at a given oil-wheat exchange rate. So, for instance, the wheat barons might be willing to let 150 million bushels go for five bushels a barrel, but only 50 million bushels at seven bushels a barrel. Again, this implies that 100 million bushels are held by those who would give more than five bushels for a barrel, but not more than seven.

Our assumption of rationality tells us that more wheat is preferred to less wheat, and more oil to less oil—whether you are a mogul or a baron. Therefore, we know that the supply curve slopes up (more wheat per oil, more oil supplied), and the demand curve slopes down (more wheat per oil, less oil demanded).

And—in case it isn’t obvious—we see that the market clears where the curves intersect. There is a single oil-wheat ratio at which, after a single set of trades, all barons and moguls are satisfied and no further exchanges will be made. (Except, of course, as actors and preferences change.)

Pretty much everyone understands “supply and demand” in this sense. However, the critical point is that the market reveals preferences only at the clearing rate. It can tell us, for instance, that the market clears at six bushels a barrel—at which point the moguls are collectively willing to trade 20 million barrels of oil for 120 million bushels of wheat, and the barons collectively willing to trade 120 million bushels of wheat for 20 million barrels of oil.

The market does not reveal the shape of either curve, except at the intersection of the two. It does not reveal how many barrels the moguls would release at five bushels a barrel, or four, or seven. It does not reveal how many barrels the barons would want at five, or four, or seven. This is one of the most important insights of 19th-century economics—now you know where this blog got its name.

Great. Now, we understand markets. So clearly, for every two commodities in our economy—oil and wheat, or silver and wheat, or wheat and corn, or oil and silver—there needs to be a market. Since obviously, anyone can want to trade anything for anything else.

Actually, this is not true. If we have \( n \) commodities, we don’t need \( O(n^2) \) markets. We only need \( O(n) \) markets. Whew! What a relief. You’ll note that in real life, there is no oil-wheat market—even though farmers drive tractors, and roughnecks eat bread.

Imagine a square matrix in which our \( n \) commodities are rows and columns. How much real information is there in this matrix? Obviously, the oil-oil exchange rate is always 1, and the oil–wheat and wheat–oil rates are the same thing. So half our boxes, plus \( n \), become blank.

But this is by no means enough blanking. Because we note an interesting fact—if this entire marketplace clears, it cannot be possible to construct a cycle of exchanges through which one ends up with more wheat, or oil, or silver, or anything, than one started with. This is obviously a desirable trade, and yet in a cleared market there are no desirable trades.

So the matrix is quite degenerate. There are only \( n-1 \) meaningful values. For instance, if we know the exchange rate of oil to wheat, silver to wheat, and corn to wheat, we can trivially work out the exchange rate of corn to silver. Wheat, in this example, is our numéraire. In theory, any commodity can serve this role.

You might say: well, in that case, why wheat? Why not silver? Indeed silver seems, just intuitively, like a much better denominator than wheat. Of course the commodity generally used as numéraire is that generally known as “money,” and of course silver has served many societies as “money.” And an exchange rate for which “money” is the denominator is known, of course, as a price.

Note how far we have drifted from our medieval urge to objective value. Before this little indoctrination session, you might have said your house was worth $400,000. Now, you say that you feel confident in finding some party B who would exchange 400,000 dollars for your house—but much less confident in finding a B who would fork over, say, 450K.

This is, of course, a case of hypothetical and unrevealed preference, until you actually sell the freakin’ house. Then, if you actually get 400, you may feel justified in describing this as its price. (But then it’s no longer your house.) The word price is dangerous—you can feel it wanting to drop its denominator, and become a unitless objective value. But it is impossible to avoid. One can only exercise the utmost caution in its presence.

We could easily be satisfied in this definition of “money”—the standard denominator for transactions. But the scare quotes remain. Because this definition is quite unsatisfactory.

Again, the standard denominator of exchange is arbitrary—the numéraire could just as well be wheat as silver. But somehow, in reality, it is often silver, and almost never wheat. Why is this?

Indeed the numéraire is arbitrary, especially with modern electronic accounting. It is a cosmetic role. We could indeed all do our books in wheat. But the intuitive assumption—that whatever is the standard denominator for transactions, the numéraire, becomes “money”—is quite wrong. The causality goes the other way around. Money becomes the numéraire. There’s something going on here, Mr. Jones.

The cosmetic role hides a much more important phenomenon: the anomalous demand for “money.” Intuitively, the commodities historically used as “money,” such as gold and silver, are not particularly useful to anyone. Yet they are highly sought after. Carl Menger, whose theories on the subject are almost but not quite right, put it like this:

There is a phenomenon which has from of old and in a peculiar degree attracted the attention of social philosophers and practical economists, the fact of certain commodities (these being in advanced civilizations coined pieces of gold and silver, together subsequently with documents representing those coins) becoming universally acceptable media of exchange. It is obvious even to the most ordinary intelligence, that a commodity should be given up by its owner in exchange for another more useful to him. But that every economic unit in a nation should be ready to exchange his goods for little metal disks apparently useless as such, or for documents representing the latter, is a procedure so opposed to the ordinary course of things, that we cannot well wonder if even a distinguished thinker like Savigny finds it downright ‘mysterious.’

(Prof. Dr. Menger declined to comment on the still more advanced civilizations of the 20th century—which managed to dispense with even the metal disks. “Progress,” he muttered, and slammed his portcullis on your correspondent’s toe. Inside, someone cocked an arquebus.)

This anomaly provides our definition of “money.” By UR’s house definition, any commodity which experiences Menger’s anomalous demand is “money.” By this definition, if there is anomalous demand for silver, but none for wheat, silver is “money” and wheat is not. But we can still construct a marketplace in which silver is money, and wheat is the numéraire. It would certainly be weird—but it would work perfectly.

Of course one can define this big, vague word, “money,” however one likes. If you define it as the numéraire, wheat is “money.” But this would leave Menger’s anomaly without a name.

First we must discard one plausible theory of the anomalous demand, which is that money’s use as numéraire causes the anomaly (rather than vice versa). In our abstract marketplace, this hypothesis is easy to falsify.

If you wish to trade oil for corn in a wheat-priced market, you do indeed need to sell oil to buy wheat, and buy corn with that wheat. So you need wheat. However, your demand for wheat is entirely transient—it exists only for the time it takes your trade to execute. Thus the total stockpile of wheat required to lubricate a wheat-priced market is the maximum amount of wheat that concurrently executing trades will use in parallel. On an efficient electronic market, this number will be negligible. On a really efficient market, it will be zero.

Thus, the use of wheat pricing in a marketplace creates no significant demand for wheat in that marketplace. Thus, it cannot possibly be the cause of any such anomalous demand. Similarly, if there is anomalous demand for silver, we know it cannot result from the use of silver as a numéraire. Thus the causality can only run in the opposite direction. First, silver experiences anomalous demand; then, it becomes the standard denominator.

The actual cause of the anomalous demand is no big mystery. It is the natural desire of many actors to make exchanges across time.

Rather than exchanging present oil for present wheat, an actor may wish to exchange present oil for future wheat—for instance, if he has oil now and wants to make sure he has bread next year. (In some cases this demand can be satisfied by a commodity futures market, but this requires the actor to know exactly what good he wants for his oil, and when. Certainty about one’s future demand schedule is certainly not an implication of rationality.)

Therefore, actors engage in a behavior for which the natural English word is “saving.” But this word was abused to well past an inch of its life by the 20th-century economist. To be precise, therefore, I will call it “caching.”

In caching, actor X exchanges A at time T1, for B at time T2, with two exchanges. At T1, he trades A for some caching medium M. He then stores M until T2. At T2, he trades M for B. Rocket science, this is not.

Key point: the demand is anomalous because X does not actually consume M. For instance, if M is silver, X does not make jewelry out of it, use it as a paperweight, pour it down Crassus’ throat, etc. Therefore, ceteris paribus, use in caching increases demand for M. Hence the anomaly.

If T2 equals T1, caching degenerates into the transient transaction described above. Transient transactions do not create anomalous demand. Non-transient transactions do. At any time T, we can observe a stockpile Sm—yes, the “money supply”—of this marketplace. This is simply the sum of all M stored by caching actors. It clearly represents new demand for M—and the longer the caching period \( (T2 - T1) \) of a cache, the longer it contributes to Sm.

So, for instance, if you are paid in silver and live paycheck-to-paycheck, you make only a minimal contribution to Sm, because you never hold very much silver. The boss gives you a little leather bag of it on Friday; you blow it over the weekend on Colt 45, cheeba and Doritos; which run out by Thursday; and so on. You are conducting exchanges across time—but not much time. It is your boss’s boss’s boss, the fat Armenian, with the big barrel of thalers in his basement, who really makes money money.

Thus in the case of the anomalous demand, we see one stabbed corpse and one bloody knife. And thus we feel comfortable in calling any widely used caching medium money. Moreover, since cachers of silver are inclined to do their accounts in silver, it is not hard to see how silver becomes the standard denominator of transactions—the numéraire.

This does not tell us why silver is a good caching medium, and wheat is not. But it gives us the tools to at least ask the question.

We must ask it subjectively, of course. From the perspective of actor X, wishing to exchange A at T1 for B at T2, choosing some good to serve as M, what makes some Ma better than some other Mb? Does this depend on: the nature of A? The nature of B? Or the subjective desires of X?

Surprisingly, the answer is “none of the above.” X’s choice depends only on Ma, Mb, T1 and T2. It is objective; it depends on future information, but objective future information. If the exchange rate Ma/Mb at T2 will exceed the exchange rate Ma/Mb at T1, with any storage-cost differential factored in, Ma is a better caching medium than Mb. Otherwise, Mb is better. In either case, one medium beats the other for any A or B.

In plain English, if Ma is appreciating against Mb across T1-T2, Ma is a more desirable money than Mb. Otherwise, vice versa. Of course, this calculation must include the cost of securely storing Ma or Mb for T2-T1, which is prohibitive for many goods—salmon, for instance.

Of course, this requires X to know the future of two commodity prices (Ma and Mb). This is by no means a trivial feat of prophecy—highly profitable, indeed, if it can be done. But prophecy is in fact the least of X’s problems. He has a worse problem. He is in the land of game theory.

The trouble is that X is just one actor in a world of many other actors. And because any widespread use of any Mx as money increases the demand to exchange any A for Mx, ceteris paribus, it will also increase the exchange rate between A and Mx. Moreover, if many actors use Ma for money but not Mb, the exchange rate Ma/Mb will increase.

But wait—this is the input to our calculation. Unfortunately, it also seems to be the output. The computation is self-referential. X is looking for a Nash equilibrium—he needs a strategy that works well for him, assuming everyone else is using the same strategy. I.e.: game theory.

Fortunately, good mathematics never conflicts with good sense. So let us confine ourselves to the latter, and apply it to our silver example. Starting from an imaginary prehistoric point in which there is no money at all, suppose every X who produces anything—wheat, oil, salmon, frozen concentrated orange juice, any A—chooses to cache in silver, exchanging these products for silver.

Ceteris paribus, this new demand decreases the exchange rate from wheat to silver, oil to silver, and anything else to silver. I.e., it raises the price of silver in wheat, oil, and anything else—including other potential Mx, such as gold. Silver, previously a minor industrial metal of minimal utility—pace Menger—experiences a rush of anomalous monetary demand which greatly exceeds its previous industrial demand.

Thus, silver’s price in wheat, silver’s price in oil, etc., etc., all skyrocket. But most important to the cacher, silver’s price in the alternative monetary good—gold—skyrockets as well. Thus if Ma is silver and Mb is gold, Ma/Mb skyrockets—since everyone is caching silver, and no one is caching gold. So, assuming that everyone is caching silver, everyone should cache silver. And this is our Nash equilibrium.

On the other hand, this logic is not metal-specific. It works exactly the same way for gold. If everyone is caching gold and no one is caching silver, everyone should cache gold and no one should cache silver.

And in the worst case for X, at T1 everyone is caching silver. Therefore, at T1, X trades his wheat for silver, buying silver at the stratospheric price created by all other silver cachers. Sadly, for reasons unknown, somewhere between T1 and T2, everyone decides to exchange their silver for gold. They shouldn’t, but that doesn’t mean they won’t. Sadly, X is the last to hear of this trend.

When all the formerly cached monetary silver returns to the industrial silver market, we have the reverse of anomalous demand—anomalous supply. Thus, ceteris paribus, we would expect the price of silver in anything to be depressed below its normal equilibrium. Of course the price of silver in gold will suffer the worst, for gold is going up as silver is going down. So X buys high, sells low, and then has to buy high again. Yea, truly is he jobbed in every orifice.

So what M should you cache? Confused enough yet? Let’s take a step back and try to profit from this confusion.

First, we see that at least one good must experience anomalous demand and become money, because the pattern of caching is a human universal.

Second, we see that the problem is path dependent. We have yet to establish the exact criteria that make a good M, but obviously multiple goods (silver and gold, in our example) can satisfy these criteria. Yet it is perfectly possible to construct an economy in which actors cache in silver but not gold, or in gold but not silver.

Third, we see that coexisting monies are unstable. A natural question is: given the uncertainty, why not cache in both gold and silver? Why doesn’t everyone just diversify their portfolio, so to speak? This seems much more stable than picking one of the two. In fact, it is much less.

The problem is that anomalous demand is a positive feedback loop. Suppose both gold and silver are monetized. But they are separate goods, each with its own supply and demand, so the bimetallic ratio, Mg/Ms, cannot be fixed. And once it starts to drift in favor of either, cachers will recognize that drift and convert their caches to the money which is gaining. Causing it to gain even more—positive feedback. Eventually, one of these goods will be monetized and the other will lose its anomalous demand. After suffering from anomalous supply, of course, while industry works off the monetary stockpile.

(This actually happened, against silver and in favor of gold, in the late 19th century. Gold and silver had coexisted as money since classical times, largely because the world was an inefficient patchwork of gold countries and silver countries—China and India, for instance, were on silver. As trade became global and efficient under British hegemony, Britain being on the gold standard, silver could no longer compete and was effectively demonetized.)

No matter how many goods you try to diversify your anomalous demand across, in an efficient market this “Highlander effect” will leap up and bite you in the butt. Since X is seeking a Nash equilibrium, he must assume that if he is diversifying, others are following exactly the same strategy. Thus they are smearing their anomalous demand across a basket of goods, jacking up the price of each a little rather than one a lot. Nonetheless, what can be jacked up a little can also fall back a little—and will, as anomalous demand concentrates in a single good, with the first to buy that good profiting the most. “There can be only one.”

And thus the problem is revealed as not a problem at all—under normal circumstances. Under normal circumstances, the good to cache is the good that everyone else is caching, i.e., money. Under normal circumstances, no one sees the origin of money—money is already there. All the anomalous demand is concentrated into a single good. That good is money, and no grasp of game theory is required to tell what good it is. If a time machine transported you into any normal, sound economy, it would be immediately obvious.

Under normal circumstances, it is extremely imprudent to cache in any other good than the standard money of the economy—e.g., to cache in gold, when the standard is silver. Why? Because if you are doing it, other people are probably doing it, too. You are probably not the first. Thus you are paying a monetary premium for your gold, and thus caches already exist. If more people buy in after you, if more anomalous demand appears, the premium will increase and you will profit. But if at any point this trend reverses, it has nowhere to go but down, and you will get jobbed as described above. You can only win if gold takes silver’s head and emerges as the new monetary standard.

This pattern may seem familiar to you. There is actually a word in English for an abortive attempt to create stable anomalous demand. That word is “bubble.” Anything—a metal, a baseball card, a worthless Internet stock—will keep going up if people keep buying in. Anomalous demand. Also referred to as the greater fool theory.

But there only has to be one good which experiences stable anomalous demand. That good is money. Money is the bubble that doesn’t have to pop. Any good that tries to replicate this stability alongside the current monetary standard will either (a) replace that standard, or (b) pop. Probably the latter. If your Internet stock does not actually have a chance of becoming the new global currency, therefore, you are well advised to price it as if no anomalous demand existed—and avoid the greater fool theory.

Intuitively, you can think of a monetary system as a sort of battery. When actors cache in silver, they are charging the silver battery—pressurizing the bubble that doesn’t have to pop. When they spend their caches, they are discharging the battery. Naturally, the exchange rate between wheat and silver is set by the collective desire to exchange silver for wheat—discharging the battery—and the collective desire to exchange wheat for silver—charging the battery.

This analogy breaks down in one important place: there are no units with which we can measure the “charge” of the battery. All we see is the set of exchange rates between other goods and silver. Each of these exchange rates has the same denominator—silver—and a different numerator—oil, wheat, etc. We can no more mix them in calculation than we can add furlongs to bushels. Broadly, we can see that silver must be money, because we observe Menger’s anomaly—its desirability is completely out of whack with its utility. But there is no precise means by which we can quantify this anomaly.

In practice, however, we see that the anomalous demand for a standard monetary good is so great that the normal, industrial demand appears negligible. It can even be zero, although it cannot have always been zero—or no one would have cached the good in the first place, worthless objects being worthless, and it would not have become money. However, once the good is standardized as money, it is stable and can be as useless as it likes.

If we assume for purposes of argument that anomalous demand is the only demand for money, we observe an interesting phenomenon first noted by Hume. Since money is not used by those who cache it, a supernatural force which replaces every gram of silver with two grams—assuming said force rewrites all contracts which mention silver accordingly—will leave the economy exactly the same. Everyone who, yesterday, was willing to exchange a bushel of wheat for a gram of silver, will today demand two grams of silver. The charge in the battery is unchanged; its size is doubled and its density is halved.

But note that this assumes an identical distribution of the silver. If our supernatural force doubles the amount of silver in the world by quadrupling the caches of Americans, while leaving the caches of the unfortunate Europeans alone, we will see increases in the exchange rate of pickup trucks to silver, but not in the exchange rate of escargot to silver. As for goods that both desire, the Americans will get more and the Europeans less. This is known as the Cantillon effect. Our simplistic battery analogy has no room for it.

Likewise, our supernatural force can create a zillion tons of silver without affecting prices at all, if those zillion tons are placed in the custody of an actor who does not spend them (i.e., whose demand for any good is zero at the present price). Again, our battery analogy has no room for this weird corner case. The lesson: use the battery analogy, but be aware of its limitations.

We are now in a position to understand the qualities of a stable monetary commodity. First and foremost, money must hold a charge—it must respond to anomalous demand with a stable increase in its price vector. It must not respond to anomalous demand by a mere increase in production.

Thus we see the difference between silver and wheat as monetary candidates. Wheat is a fungible and storable commodity—it has storage issues, but let us disregard these for a moment. It is not transportable, but electronic warehouse receipts can surmount this. The problem with wheat as money is that no amount of anomalous demand can increase its price above the cost of farming, which can generate an indefinite amount of wheat.

In the battery analogy, wheat leaks. Suppose that the world is on a silver standard, when suddenly our supernatural force destroys all silver and removes the very element from the periodic table. The market must choose some new money. If some set of actors try to treat wheat as money and buy into a wheat bubble, they will not send the price of wheat to the moon as they build up their caches. Rather, they will generate an arbitrarily absurd stockpile of wheat. Meanwhile, those who bought gold instead will be sitting on their profits. The wheat bubble will pop, the wheat stockpile will be sold as food, the wheat cachers will get jobbed.

Compare this to a precious metal, such as gold or silver. These are monetary metals precisely because of their restricted supply. Mining precious metals is an exercise in diminishing returns—the more you extract, the more expensive it becomes to extract more.

If our present accursed monetary system were to vanish from the earth, the ideal monetary replacement would be gold rather than silver, simply because annual gold production (leakage) is only about 3% of the global monetary stockpile (which is large, because gold retains a nontrival monetary role.) This is very significant leakage, but it is much less than the leakage in silver—in which annual production is comparable to the global stockpile. Thus it is much harder to monetize silver, because early buyers will be diluted by a much larger wave of new supply.

Since we understand supply and demand, we can see that any new money produced satisfies and thus cancels our anomalous demand. Since we understand that the anomalous demand for money is self-dependent, we see why people take leakage so seriously. Indeed there is another name for monetary leakage—“counterfeiting.” On a natural metal standard, mining of metal is the precise equivalent of counterfeiting.

Leakage is extremely dangerous not just because it effectively equates to mass confiscation, but also because it destabilizes the monetary standard itself. Assuming, as in Hume’s example, that leakage is uniformly distributed (which it won’t be), we would expect to see a natural upward drift in all prices as anomalous demand is diluted. Note that this includes prices of potential competing moneys—Mb to the present standard Ma. If silver leaks, wheat prices in silver will tend rise—but so will gold prices in silver. (Assuming that gold does not have its own issues.)

And when Mb/Ma is increasing, we know what this tells us. It tells us that gold is a better metal to cache than silver. If this trend can run to completion, replacing silver with gold as a monetary standard, it may well do so. And there is no reason it has to do so slowly. The result will be a monumental economic cataclysm in which wealth is redistributed from the aforementioned argentiferous Armenians to a new generation of aurophilic speculators—almost certainly Jewish. Few political systems can withstand such an event, good for the Jews though it is.

In an ideal monetary standard, the stockpile of money would be fixed, and there would be no mining or counterfeiting whatsoever—and thus no leakage. No natural good satisfies this criterion, but artificial commodities can be constructed—i.e., “fiat currencies.” Thus we see another counterintuitive truth: a fiat currency can in fact be a better money than gold or silver. Unfortunately, it also lends itself much more readily to pathological abuse. Paul Erdös became a world-class mathematician by taking speed; you or I would just become speedfreaks.

We now understand money. Which means we’re almost done! All we need to understand is interest and the loan market.

Loans are extremely simple. A loan is a promise of future money. Specifically, it is the promise, written by some party Y, to pay some quantity Q at some future date D. The price of this loan is the price of future money in present money, modulo the probability that Y is a scumbag or loser and will default on the loan.

Default risk is easy to factor out of this equation, so let’s just deal with it now. How does a market estimate the probability that a loan will default? This probability itself can be isolated and sold, via an instrument such as the (unjustly) notorious credit-default swap. Thus, the price of a risky loan is the price of a risk-free loan, minus the price of a risk-free CDS on said loan.

A CDS market is a case of a prediction market. How do prediction markets predict the future? Magic? No, supply and demand. If you think the price of a prediction instrument does not match its actual probability of payout, and you are right, you profit. If you are wrong, you lose. There is no magical law which requires the intersection of supply and demand to equal the actual probability—in fact, defining “actual probability” is itself an almost magical challenge.

Thus, the price of a prediction instrument is only as good as the collective intelligence of its participants. Successful prediction markets, such as Vegas for sports, are very hard to beat, because they have been active for a long time and experienced a Darwinian effect—gamblers who know their sports win and keep playing, gamblers who don’t lose and drop out. Exactly the same effect is seen in any non-dysfunctional financial market.

Even without these exotic instruments, loans are easy to standardize by estimating default risk and constructing pools of diversified loans, thus putting the central limit theorem on your side. One loan with a 5% chance of default is a very scary thing to buy, but an equivalent share in a pool of 100 uncorrelated loans, each with the same chance of default, is much easier to handle. Of course, the probability must still be estimated, but this is a well-known profession—estimating the default probability of loans is (in a sound economy) the job of “banking.” Like any job, it’s not easy but it can be done.

With that said, let us analyze the loan market assuming zero default risk. Again, the price of a loan is the price of future money in present money. This is always less than 1—e.g., a 2011 gram of silver might cost 900 milligrams in 2010 silver.

But what does this mean? Why does this market exist? Let’s look at the demand and supply sides separately.

Consider our cacher X again. Under what circumstances should X buy a loan? Answer: X should buy a loan if, and only if, he knows his T2—the date at which he will spend his cache. If his T1 is 2010 and his T2 is 2011, he will have more silver in 2011 if he buys 2011 silver with his 2010 silver, rather than just caching his 2010 silver until it turns into 2011 silver.

But wait—is this leakage? Does the existence of a loan market decrease anomalous demand? Not at all. The silver cache has just been transferred—instead of X holding the silver, it is Y, the borrower. Neither anomalous demand nor the silver stockpile has changed in the slightest.

A more interesting question is why X should not buy a loan if he does not know his T2. He has a barrel of silver in the basement—he might want to spend it on something in 2010, or he might want to spend it on something in 2011. Suppose he just buys a 2011 loan, anyway? A loan is a negotiable instrument—anyone can resell it. If X decides that he wants to spend his cache in 2010, he can just sell the loan on the open market, get the money back, and spend it.

As it happens, we have already solved this problem. We solved it above in our discussion of anomalous demand. X, when he performs this unnatural act, is caching a good (the loan) which he may, or may not, use as money—buying and reselling, as above. If he turns out to actually want 2011 money, he has correctly anticipated his future desires and is not using the loan as money. Otherwise, his demand is anomalous. If he is the only one doing it, fine—his actions will have a minimal effect on the market. But he won’t be.

Any societal pattern of such behavior will create exactly the same bubble phenomenon that we see in any failed attempt to create an alternative money. When the herd of Xs, who should simply be caching money, buy 2011 loans, they drive the price of 2011 loans up (lowering interest rates). When they sell 2011 loans—because they actually wanted 2010 money, not 2011 money—they drive the price of 2011 money back down. Therefore, unless they get into the bubble early and out early, they get jobbed. Just as in any bubble. Lesson: don’t buy any good you don’t intend to use, unless that good is money.

We also see that the longer the term of the loan, the lower the market demand for it. For any X, two one-year loans, strung back to back, is an adequate substitute for one two-year loan. But for an X whose T2 is in 2011, not 2012, only a one-year loan will suffice. Thus the longer the loan, the fewer buyers who can use it.

In real societies, however, demand for long-term loans is quite nontrivial. If you intend to retire 20 years from now, it is quite reasonable to buy a 20-year loan to fund your retirement. As we’ll see in a moment, you will get a better interest rate than if you strung out 20 one-year loans in a row.

Now, let’s consider the supply of loans. Who would sell a loan? There are many reasons to borrow, of course, but generally the borrower is involved in some productive endeavor. Your endeavor certainly has to be productive if you’re getting 900 kilos of silver in 2010, and turning it into a metric ton in 2011.

On a precious-metal standard, the simplest form of productive enterprise is a mine. You spend your 900 kilos digging the hole; you pull a metric ton out of the hole. A business that earns its return through commerce, rather than monetary production, looks no different to the lender.

The important point about productive enterprise is that production always takes time. There is no possible enterprise that can turn 900 kilos of silver into a ton of silver in one second. This dictates the term of the loan.

And again we see that the supply of 2-year loans will exceed the supply of 1-year loans at any interest rate—because an enterprise that produces return after 1 year can sell a 2-year loan by selling one loan after another, whereas one that actually takes 2 years is quite ill-advised to sell a 1-year loan. (The result being the inverse of the bubble scenario described above.) If you ever see an Austrian economist talking about “more roundabout means of production,” all he means is this.

When we combine these supply and demand functions, we learn something very important about the sound economy: that its yield curve (the market interest rate for every term) slopes upward.

Interest rates at longer terms are higher than those at lower terms, because loan prices are lower at longer terms; longer loan prices are lower because the demand is lower, and the supply is higher. No ceteris or paribus is necessary. If you see an inverted yield curve, you know that you are looking at a very, very sick loan market. Don’t walk away—run away.

Which brings us to our final subject: pathological economics.

Barring bizarre and unforeseen real-world phenomena, such as a zillion-ton asteroid of silver landing in Brazil, a sound economy is stable. Supply is always equal to demand, bubbles do not exist in commodities or financial instruments, and the monetary standard suffers only negligible leakage from counterfeiting or mining.

Unfortunately, misgovernment is the fate of man. Misgoverned economies appear to be an especially eternal fate. In fact, there are almost no historical cases of a sound economy as described above. Someone is always screwing up something.

There are two basic patterns of economic misgovernment. One: the government fixes prices. Two: the government tampers with the monetary system. The former is heinous; the latter is diabolical.

Fortunately, just about everyone knows the effect of price-fixing: surplus or shortage. Somehow, the principle of supply and demand has actually become part of democratic folk wisdom, like multiplication tables, the names of the Greek gods, and pressing “1” to get your voicemail. We must be thankful for this mercy, which surely cannot last forever.

The simplest way to tamper with the monetary system, known since well before Jesus was a little boy, is to debase a metallic currency. The state simply forces its subjects to accept the debased currency at par with the old, good money—e.g., to accept an alloy of silver and nickel in payment for debts in pure silver. Thus it creates leakage, without actually creating silver. It taps the battery.

In many cases, debasement is more convenient than taxation, to which it is obviously equivalent. Some remarkably effective governments have employed it. For instance, during the Seven Years’ War, Frederick the Great (fighting for Prussia’s life against a coalition of all the surrounding countries) debased the currency. Of course, he also restored it within a year after the war. That’s because he was a king, not a professor. But I digress.

Far nastier is the practice of debasing the currency with loans. Thus, rather than being compelled to accept a silver-nickel alloy as silver, subjects are compelled to accept a promise of future silver (typically written by the government or its closer friends) as present silver. The output of this leakage is not silver, but lower interest rates. The practice is especially insidious because its blessings are visible to all borrowers, whereas the curse of monetary dilution is almost invisible.

Modern fiat currencies were not, like the optimal fixed-supply fiat currencies I would like to see, created all at once by actual fiat. They were created by incremental loan debasement. The 19th-century classical gold standard is quite misnamed, for instance—it was not actually a gold standard, but a pound standard. Behind the pound sterling was a certain amount of gold, and a certain amount of British national debt—denominated in gold. Not at all the same thing.

The pound, of course, was “convertible” into gold. Similarly, shares in Bernie Madoff’s Ponzi scheme were convertible into dollars. Many people converted them into dollars, and felt convinced as a result that they were “worth” dollars. Of course they were simply shares in the underlying assets, and could be worth no more or no less.

Thus a pound was a share in a pool of gold, and a pool of British debt—the latter larger than the former. After WWI, the latter became grossly larger than the former; the pretence could no longer be sustained; and Britain “went off gold.” In reality it had not been “on gold” since the 17th century. It had been running a hybrid currency—part gold and part fiat. The fiat quality arose from the fact that Britain’s national debt, though denominated in gold, could not be paid in gold. Thus in gold terms these were simply bad loans, i.e., worthless paper—i.e., fiat.

Normally, in a bankruptcy, debt becomes equity, and so the old gold-denominated government debt morphed (without any formal process) into its modern state of government equity. And thus we see the modern fiat currency, a unit of which is essentially a share in the government. Without voting rights or dividends, alas, but a share nonetheless. Equity is the lowest tier of debt. If a note confers no rights, it is equity.

On an (open-ended) fiat currency the potential for pathological finance is almost unlimited. For instance, perhaps the most pernicious practice of fiat finance is the issuing of loan guarantees, which now in USD total in the tens of trillions. A loan guarantee is simply a camouflaged loan, disguised for the usual corrupt reasons. When the government guarantees A’s loan to B, what is really happening is that A is lending to the government and the government is lending to B. Thus, for instance, when FDIC guarantees your loan to your bank (a “bank deposit”), you being A and the bank being B, you are lending to USG (whose payment is sure, because the loan is denominated in USG’s own equity), and USG is lending to the bank. The latter is the corrupt transaction; the former is the disguise.

But we could go on in this vein forever. Again, the point to repeat: there is no simple or gradual way to unwind a pathological financial system. It must either collapse or be destroyed. On the other hand, there is no reason it can’t survive forever, like a Cuban truck—stinking horribly, bellowing like a bear giving birth, and running no faster than fifteen miles an hour. All we can hope is that, however it turns out, it will be good for the Jews.