Okay, once again I’m going to sacrifice my body here, risk looking stupid by asking what seems to me to be a vexatious question. Here’s the setup:
When you exchange some of your money (bank deposits) for some shares of Apple stock, those shares aren’t removed from the supply of Apple shares. (Likewise your “money”; it still exists.) The stock-supply, at least, is unchanged.
When you buy some apples for consumption, they are removed from the supply — both stock and flow.
Can we model or think about these two markets in the same way? One is a circular flow in which supply is never consumed (that is the sine qua non of financial assets: they embody exchange value that can’t be consumed to derive human utility). The other is a conveyor belt, where the supply falls off the end and disappears (or magically transforms into “utility” via consumption), with real production/resource constraints at the beginning.
Related: Clower/Burshaw on the difference between “stock supply” and “flow supply.” Or peruse the literature here.
I’ve wrestled with this before, as have my thoughtful commenters therein.
But nobody has ever come back to me with thoughtful discussion of stock supply and flow supply, or satisfactorily answered the question at the end of that post (accompanied by a Holy Grail clip that is a propos). The question posed above leaves me even more perplexed.
Am I foolish to suggest that the central concepts of economics, supply and demand, are embarassingly un(der)theorized?
There’s a great example in this Felix Salmon post, discussing the difference between the global “supply” and “demand” for bonds — which here seems to mean “bonds issued” and “bonds purchased” (designated in dollars). Mustn’t these values as implicitly defined here be identical — at least when viewed ex-post — making a discussion of their difference conceptually problematic?
Cross-posted at Angry Bear.
Comments
15 responses to ““Supply” and “Demand” for Financial Assets”
Start with a fixed stock of land. P is the price of land.
Stock: P adjusts until the actual stock of land is willingly held. The stock demanded equals the actual stock.
Flow: some people want to sell a flow of land per year, and others want to buy a flow of land per year. P adjusts until flow supply equals flow demand.
Now suppose the Dutch will produce a flow of new land, at a price. That adds to flow supply, and means the actual stock is growing over time.
Now suppose miners consume a flow of existing land, at a price. That adds to flow demand, and means the actual stock is falling over time.
Apples are like land, with lots of Dutch and lots of miners.
“Can we model or think about these two markets in the same way? ”
Yes by economists from James Tobin to Wynne Godley/Marc Lavoie.
You have a copy of Godley/Lavoie right – check the section on portfolio/allocation.
@Ramanan
Of course I should say “the same way” isn’t accurate but that’s what you may be looking for.
Also note you have to keep distinction between primary products and products and services of the “non-farm” sector. So we have a fix-price sector and a flex-price sector.
The market for apples is different from the market for Apple iPad where the producer sets the price.
“which here seems to mean “bonds issued†and “bonds purchased†(designated in dollars). Mustn’t these values as implicitly defined here be identical — at least when viewed ex-post — making a discussion of their difference conceptually problematic?”
Not really. You can asset demand functions etc and the supplies and demands are brought into equivalence by a price clearing mechanism.
A lot of heterodox people get confused because they are told that the market for the fix-price sector are not price clearing and hence they assume the same for financial markets.
@Nick Rowe
Thanks Nick. But that doesn’t really answer the question about financial assets vis-a-vis real goods, does it? (Land is tricky because it both acts like a financial asset and provides human utility.) Increased volume in the financial markets does not correlate with increased prices. Bear markets often have high volume.
Increased volume in the real goods market (increased “demand”?) does, generally, necessarily — because supply is consumed by the increased demand/volume of purchases, and production is resource-constrained.
@Asymptosis
The distinction that matters for stock/flow is not financial vs real, but durable vs non-durable. A real good (like a consumer or producer durable) can provide a flow of services over time, and we consume those services rather than the good itself.
I’m on basically the same page as Ramanan (which *sometimes* happens 😉 ).
@Ramanan
@Nick Rowe
“You can asset demand functions etc and the supplies and demands are brought into equivalence by a price clearing mechanism.”
Sure right I totally get that. But that analyst at least (and Felix along with him) is using “demand” and “supply” to mean “amount purchased” and “amount issued” (designated not in Q, but in $ — P*Q). That is decidedly *not* what “supply” and “demand” mean in my understanding.
So maybe I’m just pointing out his misunderstanding, but I still feel like it points to a much deeper conceptual problem.
Any response to Clower/Burshaw’s tentative steps at theorizing this?
Yep. And since amount purchased must necessarily be the same as amount sold (whether we measure both in Qs or in PxQ’s), there is something decidedly dodgy about his data!
You see those sort of diagrams a lot. You see “supply” and “demand” both equal to each other up to the present, and then diverging into the future. What they are perhaps trying to say (but saying badly) is that this is what would happen to quantity supplied and quantity demanded in future *if* price (or something else) did *not* adjust to equilibrate the two. So if “demand” goes above “supply”, we figure price will rise in future to bring them back into equilibrium.
None of the above was a stock/flow issue.
I skimmed Clower/Burshaw very briefly. It looked OK to me.
For instance, Wikipedia, which pretty much represents the common understanding:
“The quantity demanded is the amount of a product people are willing to buy at a certain price.”
Wrong!!! As I understand it, “demand,” properly conceived, is the whole demand curve — the amount people would be willing to buy along the whole range of possible prices.
What wikipedia seems to be describing is Q at a certain point on the demand curve.
Ex-post measurement of Q tells us nothing definitive about S and D. Especially when, as with bonds, Q can only be designated in $s, which is a P*Q measure, not (usefully) in units.
@Nick Rowe “amount purchased must necessarily be the same as amount sold ”
Right: ex-post looking back, QSupplied = QDdemanded = Qsold/Qpurchased.
But I still find it problematic when we have a Q axis and a P axis, and Q is measured/designated in P*Q…
“there is something decidedly dodgy about his data!”
I’d say, likewise, there’s something decidedly dodgy about his understanding of supply and demand.
@Asymptosis
Wikipedia is correct. “Quantity demanded” is a point on the demand curve, and “demand” is the whole curve.
Nope. Those 3 quantities are only equal at market clearing equilibrium. If there is a binding price floor, for example, quantity supplied > quantity demanded = quantity sold/bought
SR,
Suppose the US Treasury were to announce an auction of $100bn where you pay $100 for each bond at the beginning and receive $20 every year and $100 at the end for each bond, there will be a huge riot to get the bonds. So “high demand”.
If it instead says you pay me $100 now and receive $100 at the end of 10 years for each bond and no coupon at all, nobody buys it. So demand is zero.
So demand is a function of price i.e, D(p) or equivalently D(coupon) – equivalent not equal.
Instead it will hold an auction and clear the market via prices. In this case it is coupon but conceptually equivalent. The coupon is such which solves the equation
D(coupon) = $100bn.
So suppose the auction result came out with a result coupon of 3.
That means D(3) = $100bn.
Suppose on the same day, Apple also holds an auction to issue bonds.
Some people may find Apple bonds more attractive.
So D for Treasuries has now become more complicated.
Suppose the one without Apple bonds cleared at a coupon of 3.
In the presence of Apple bond it will clear at maybe 3.1
You can introduce more complications – suppose on the day investors feared US default. This may be described as a shift in the demand curve and hence the auction may clear at a coupon of maybe $4. (ie the US government would have a pay a higher coupon than otherwise).
Suppose on the day there is no fear and suppose Google purchases its own stock. Investors who have more money instead of stocks may be get interested in the US Treasury auction and if that is so the auction may clear at a lower coupon – say $2.9.
Suppose on the day banks raise deposit rates without the Fed actually raising rates. Some investors may say “you know what, I am no longer interested in the US Treasury auction since I am happy earning the interest on deposits”. So the auction may clear at a higher coupon rate such as $3.3.
Suppose on the day there is a rumour that the Fed is going to hike rates soon. Since long term rates are partly expectations of short rates, the demand curve will shift at the auction will clear at a higher coupon.
All this is in the primary markets. The whole market is bigger because of secondary market also.
Sometimes people say the statement “demand for US Treasuries fell as demand for stocks rose” silly because for every buyer there is a seller in either. But this reason itself is silly.
What it means is that the demand functions have moved so that the markets clear bonds at lower prices than yesterday at stock at higher prices.
The thing becomes conceptually complicated because in simple examples one has simplified supply-demand diagrams. But I suggest you look into the literature and find ones which have less diagrams.
So let me give an example.
Normally the value of a point in the demand curve will be higher at higher yields for government bonds. But suppose there is a fear of default – such as Greece. This will loosely speaking “reduce the demand” for Greece bonds.
But you sayz “supply = demand”
What technically it means is that the demand function is a function not just of the yield but the expected return which depends on expectations. It will then clear at a lower price – the price point where expectations have reversed. Of course expectations needn’t be right so the next day it may clear at an even lower price.
But this may be difficult to represent in a diagram.
The thing is supply-demand is not wrong. It is how one misuses it which is wrong.
Going by your post and comments, I get the impression that you find something horribly wrong about supply-demand but IMO that isn’t the case. It’s just that sometimes people misuse it and other times people misunderstand a right analysis.
Going back to your original setup, and pruning off the “stock and flow” analysis, which can come later, we have this:
When you exchange some of your money (bank deposits) for some shares of Apple stock, those shares aren’t removed from the supply of Apple shares. (Likewise your “moneyâ€; it still exists.)
When you buy some apples for consumption, they are removed from the supply
That seems to me the crux.
When you buy apples to eat, they are permanently removed from the marketplace.
When you buy Apple stock to trade, they are not removed from the marketplace.
But when you buy Apple stock to hold, they are permanently (i.e. exceptionally persistent) removed from the marketplace.
If you bury a jug of coins under the apple tree, you are hoarding the coins and they are again “permanently” removed from the marketplace.
Unlike the apples that you eat, these things you hoard may come back into the marketplace eventually. But until they do, they have an effect the same or similar to that of permanent removal.
There is a related problem with the growth of savings.
@Nick and @Ramadan:
This year’s apple harvest is 100,000 apples.
Apple IPOs 100,000 shares
All the apples and shares are sold. The apples are consumed.
The supply of apples is 0, at any price point.
The (stock) supply of Apple shares is 100,000, subject to market pricing.
Can we use the same S/D analysis to make statements or predictions about future prices in these two markets?
In the market for shares, are we actually talking about the “supply” of willing sellers (at different price points), as opposed to the supply of shares?