Wolf vs. Krugman: Separating the Economist From the (Bad) Accountant
Not surprisingly, both my allies and critics seem to be missing the point of my previous post on Krugman and (alleged) accounting identities. So some clarification is in order.
First, here’s a simple one I made in the comments on the definition of income:
If I do a consulting project for a guy and he gives me $20,000, that is clearly income. (We could come up with some weird scenarios where it might not be–like if I’m consulting for the KKK and by [so] doing I destroy my career and future prospects for income. But leave that aside.) But if my brother owed me $20,000 and paid it back, that’s not income. Just like, if I sold my car for $20,000, that’s not income, in an accounting sense.
If it were, then the “national income” of the US would be a gajillion dollars. Think of how much “income” is earned every day in the stock market from people selling shares.
OK, so to explain the relevance of the above definition: When someone pays down a debt, and the creditor receives an extra influx of cash, that isn’t how we free-market guys need to plug the hole. If the creditor just sits on the cash, then nominal income has indeed fallen. (That could be fine. Prices tend to fall too. I’m not saying there’s a problem with falling nominal incomes.)
Now if the creditor spends the money on consumption, then the creditor has dissaved to perfectly offset the saving of the debtor. Thus Krugman et al. will say in the aggregate, there has been no net saving. (The creditor’s assets will have fallen by as much as the debtor’s debt.)
What if the creditor lends out the money to somebody else? Now we’ve just replaced the original debtor’s debt, with some new debtor’s debt. Krugman wins again.
But now the point of my original article: If the creditor takes the influx of cash and buys new shares of corporate stock (or pays someone directly to make his own business better, thus investing the money), and if the corporation invests the money by paying someone an income to somehow enhance the wealth of the corporation, then (a) total nominal incomes don’t need to fall–and can even rise, and (b) aggregate debt really has gone down.
In the comments, Mammoth at least understands the rules of the chess game we are playing here, but doesn’t think it’s legal for me to Castle in this fashion. But he is wrong. Mammoth thinks that if a corporation issues more stock shares, then the previous shareholders must be the ones doing the dissaving. After all, there are now more shares of stock into which the net assets of the corporation must be divided–so haven’t the original owners’ claims been diluted?
No, because Mammoth is overlooking that the corporation now has more assets. Suppose a corporation originally has $1 million in shareholder equity, with 100,000 shares. So each share has a market price of $10. Then the corporation issues 100,000 more shares, charging $10 each. Now the corporation is sitting on its original assets, plus an additional $1 million in cash. So the corporation’s net assets are now $2 million. (Note that its liabilities haven’t gone up–shares of stock aren’t liabilities on the balance sheet.) So the corporation has $2 million in net assets, and 200,000 outstanding shares of stock. Each share thus is a residual claim to $10 in assets, just as before. (In the real world, if investors think the corporation is going to do something productive with the new capital, the share price can actually increase, even after the “dilution.”) There’s no reason to suppose that issuing new shares will reduce the net assets of the original shareholders; it’s certainly not an accounting identity.
Now for the point of my post title: If you go read Wolf’s original article–the one that prompted Krugman’s most recent display of accounting error–you’ll see that Wolf actually fills in the details of the argument. (Brad DeLong has relevant excerpts from Wolf, if you don’t want to register with FT.) It’s true, Wolf does say some things that are demonstrably wrong too, but he at least spells out the (Keynesian) assumptions behind his worldview. In context, you get the sense that Wolf realizes it’s theoretically possible for Cameron’s austerity plan to work, but Wolf thinks it’s just unlikely given businesses’ expectations, etc.
In contrast, Krugman keeps flatly asserting that his side literally rests on accounting identities. He is wrong.
But once again you seem to be saying “if we abandon the assumptions that make the paradox of thrift happen, then the paradox of thrift will not happen”.
Of course Bob! Who would argue with that?
Your counter-arguments depend on people spending.
Yes. This is Krugman’s position. If people spend, clearly we don’t have a problem.
You see this as me constantly defending Krugman – I see this as me pointing out that you aren’t an idiot and Krugman isn’t an idiot and actually you guys agree on these thought experiments you just for some reason like to argue and accuse the other of being wrong.
Which isn’t to say there AREN’T idiot Austrians, idiot Keynesians, idiot libertarians, and idiot liberals out there. There clearly are. 🙂
Daniel, then it’s not “Death by Accounting” and it’s not “the laws of arithmetic.” If Krugman said, “Death by Accounting and Controversial Keynesian Assumptions About Spending During a Liquidity Trap,” his posts wouldn’t be so smug.
OK, so Krugman is saying “under circumstances where we add two and two, the laws of arithmetic say we get four, not three”, and you seem to be saying “Krugman can’t invoke the laws of arithmetic to pooh-pooh three – after all when you add one and two that’s exactly what you get!”. I don’t think Krugman would disagree with your math. I think he’d just wonder why you’re talking about adding one and two in the middle of a liquidity trap.
Remember, Krugman is concerned about the people who claim that 2+2=3, not the people who claim that 2+1=3.
Remember, Krugman is concerned about the people who claim that 2+2=3, not the people who claim that 2+1=3.
Except I think a better analogy is, Krugman is running around saying, “The square root of 9 is +3, and anyone who says different is a liar.” Then some people are making arguments that it could be -3, and Krugman accuses them of not knowing exponentiation.
If debtors repay debt, and the creditors who are repaid spend the proceeds on consumer goods, then aggregate debt decreases.
Aggregate net worth doesn’t increase. Saving on net is unchanged. The debtors save and the creditors dissave.
Decreasing aggregate debt isn’t the same thing as increasing aggregate net worth.
Suppose every debtor pays down all of their debt to zero, and all the credtiors spend the proceeds on consumer goods. Debt is now zero. There is no debt anywhere. Can we say that debt is unchanged?
What you (Murphy) are thinking about is that the debtors pay down debt to increase their net worth, but the creditors don’t decrease their net worth either, and perhaps continue to add to it as well. How is that possible?
For aggregate net worth to increase, then investment–spending on capital goods, must increase. (This can be in real terms, and so the entire pricess could occur at lower money prices and lower nominal incomes. Still, real investment must decrease.)
In Wolf’s article, this would be a scenario where government reduces its deficit and firms continue to retain earnings but purchase new capital goods rather than purchase old capital goods or existing financial assets.
The cleanest way for this to happen would be that rather than firms using current profit retaining earnings to purchase government bonds, they buy new machines, buildings, or equipment.
Government accumulates less debt, and net worth rises because the firms are building new assets. Nominal expenditure and income are maintained. (Ceteris paribus.)
Of course, there would be many other scenarios, but if there is going to be an increase in net worth by households, new capital goods have to be produced. (As Wolf says, with the open economy, we can pay down debt to foreigners reduce the trade deficit.)
Your point about equity finance is correct too. Firms using current profit to buy capital goods _is_ equity finanice.
Still, we can imagine households paying down debt and the creditors using the funds receive to purchase stock. If those they buy stock from just feel richer and consume more, we are back to dissaving matching saving. What must happen in this scenario for aggregate net worth to increase is that some firm sells new shares to fund purchases of new capital goods.
Bill Woolsey wrote:
What you (Murphy) are thinking about is that the debtors pay down debt to increase their net worth, but the creditors don’t decrease their net worth either, and perhaps continue to add to it as well. How is that possible?
For aggregate net worth to increase, then investment–spending on capital goods, must increase. …
Still, we can imagine households paying down debt and the creditors using the funds receive to purchase stock. If those they buy stock from just feel richer and consume more, we are back to dissaving matching saving. What must happen in this scenario for aggregate net worth to increase is that some firm sells new shares to fund purchases of new capital goods.
Bill, I can’t tell if you are just (consciously) paraphrasing what I have been saying all along, or if you think you are correcting me. In any event, yes we agree entirely on these points. In my original article, and in my other examples in the comments here on the blog, I have been giving scenarios where a corporation issues new stock, and then uses the raised funds to invest back in the business.
So we have total nominal incomes not falling (in fact going up), we have aggregate debts falling, and we even have aggregate net worth increasing.There is genuine net private sector saving, even though some would have us believe this is impossible.
You are wrong Murphy. Previous shareholders are dissaving by losing ownership valued in $20K. That’s the reality and there is no way around reality.
You are making the mistake of falling into the trap of the fallacy of composition that if one share is valued $10, then shareholders owning n shares are worth n*$10. No, they are not.
If I owe you $1000 and I sell you 50% of my car valued $2K to repay my debt to you, then I am dissaving.
Previous shareholders are dissaving by losing ownership valued in $20K.
That is false. If new stock is issued for $20k, then the value of the company RISES by $20k, because the company now has $20k more cash!
So just to clarify, if you own 1 share out 100 outstanding, the company issue 100 more shares, and now you own 1 share out of 200, your equity stake remains the same? Huh, very interesting. Sorry guys, Mr. Murphy is wrong on this one. The equity ownership is absolutely diluted unless the owners participate in the secondary offering, and the funds to participate in that offering have to come from somewhere.
AP Lerner: This isn’t hard. Yes, the fraction of ownership goes down, duh. But the amount of assets being owned goes up, duh.
If you guys are right, why wouldn’t all stockholders always oppose issuing new corporate stock? According to you guys, the existing stockholders are always getting screwed, as a simple matter of accounting. And we also know that the shareholders collectively run the company. So how does that work? Every time a corporation issues new shares, it’s because there’s a majority stakeholder who wants to urinate away some of his wealth?
If the shareholders believe that the enhanced market value of the company from the additional capital raised outweighs the effect of diluted equity, then they will vote in favor of issuing additional shares.
So just to clarify, if you own 1 share out 100 outstanding, the company issue 100 more shares, and now you own 1 share out of 200, your equity stake remains the same?
No, your equity stake has fallen, but the value per share has increased by the money fetched in the equity offering.
Huh, very interesting. Sorry guys, Mr. Murphy is wrong on this one.
No, Murphy is right. YOU are wrong.
Equity is diluted yes, but if the equity sale brings in new cash, then the value per share increased (ACCOUNTING WISE ONLY, not taking into account information asymmetry and signalling theories).
If new stock is issued for $20k, then the value of the company RISES by $20k, because the company now has $20k more cash!
That is false!
Please elaborate. Are you suggesting that each additional dollar of cash obtained through the issuance is not worth an additional dollar to book value?
Exactly
And to that point, please elaborate.
Where does the lost value go, accounting wise?
Utterly, astoundingly wrong.
No, that is true!
Mammoth wrote:
You are making the mistake of falling into the trap of the fallacy of composition that if one share is valued $10, then shareholders owning n shares are worth n*$10. No, they are not.
I’m almost afraid to ask, then, Mammoth: Let’s say I own 12 shares of stock, when the market price is $10 each. How much are you saying I should value those assets?
There is no way to do it in a stock-flow consistent manner.
That’s why the stock market value is just a fiction.
Wow. No way to do accounting in a stock-flow consistent manner? I really hope this is just MamMoTh confusion and not something that all MMTers believe. Are you confusing accounting with something else? Like liquidity considerations?
Just me, not really interest in fiction.
Now I have a question. Is this assumption that diluted shares don’t lose value part of Austrian Economics or just Murphy Economics?
Mammoth wrote:
Now I have a question. Is this assumption that diluted shares don’t lose value part of Austrian Economics or just Murphy Economics?
No, this is how fractions work, Mammoth. You are saying the denominator goes up, and conclude that the quotient must be smaller. I’m pointing out that you’re missing one little step in your argument.
BTW, did you ever answer me about why the dumb shareholders would vote to dilute their own property? Corporations issue new shares all the time, so this is a huge puzzle you have to explain.
Mammoth appears to be conflating dilution of percentage ownership with a decrease in nominal wealth on the part of shareholder. In fairness, such scenarios can exist — for example, where the issuance causes a 51% holder to become a 49% holder — there is a premium attached to a controlling interest that will disappear. But, that is the exception, not the rule. Also, he seems to be making the same claim from the point of view of the company. Even if they issue shares, additional cash on the balance sheet leads to an equivalent increase in book value. He has yet to elaborate on why that is not the case.
so this is a huge puzzle you have to explain.
There is no puzzle.
Sometimes they do vote to dilute shares expecting the raised money will be invested in such a way that it would generate more income, which will be possible only if someone is willing to dissave, and the circle is closed: someone has to dissave.
Or do you think shareholders would be happy if their shares were diluted and the money raised was used to give a bonus to the CEO?
Time to admit you have more hidden assumptions than Krugman, and more unrealistic ones too.
No MF, I am just concerned about the Cantillon effects of share dilution with the injection of funny shares.
No Mammoth, each share is entitled to the same percentage of the company, which means both old and new shareholders now have a claim over the cash that has been fetched in the new equity offering.
Nobody answered my question.
This assumption that share dilution through the injection of new funny shares will never affect the share price, is it part of Austrian Economics or of Murphy Economics?
Mammoth, again, you are conflating two different ideas – market cap as determined by share price and book value. No question, following an issuance, the shares could trade anywhere. But, the incremental cash on the balance sheet, for purposes of calculating book value – it is as straightforward as you can get.
Ok, you’re going to have to spell this out. I own an asset X. The asset side of my balance sheet shows X. The liability side shows equity. Now I incorporate and issue myself one share of stock. My private balance sheet shows an asset of one stock share. My liability side shows equity (I own the share). The corporation’s balance sheet shows X as an asset. It’s liability shows the one share.
Now the corporation issues another share to raise money. I sell that one share for $100. The corporation’s balance sheet now shows assets of X and $100. The liability side shows two shares.
My own balance sheet is unchanged. Now it’s possible that the market value of a new share depends on the demand for that share. It is not functional. It could be higher that $100 if people think the corporation will create lots of value from the raised cash, it may be lower than $100, or it may be equal to $100.
But there is no functional “dilution” of my net worth due to the share issue, right? Only my ownership in some (market valued) pool of assets is changing.
I hope this makes sense. Otherwise, please use the Khan Academy to brush up on math.
I am not conflating anything, I am talking about shares.
Murphy’s example is based on share prices not falling after share dilution.
That’s possible since share prices can be whatever the last person to buy one paid for it.
But the only reasonable assumption behind that case is that shareholders expect the money raised will be invested in such a way that income of the company would rise, that is that someone is going to dissave.
Previous shareholders are always dissaving, temporarily, expecting someone else will dissave in the future.
So be it. You’re trying to be too cute by half with your argument. Whatever.
Nevertheless, your comment above that each dollar added through the issuance does not add a dollar to book value is also wrong.
Ugh. Please tell me you’re not confusing future spending with *increases* in future spending.
Suppose I plan to create a turkey farm to grow turkeys for next year’s Thanksgiving. I follow the steps in the example. I form a farm corporation. I issue new stock, etc. When people buy my shares, they need not believe that future savings will be reduced [from what they would be if I did not form my corporation]. They only need to believe that I’m going out-compete other growers as we bid for for dinner budget dollars.
Further, if stockholders in _another_ turkey corporation’s stock believe that I will outcompete them, then they may decide to sell their stock at a lower price. All of this is possible without any change in people’s spending and saving rates.
“If it were, then the “national income” of the US would be a gajillion dollars. Think of how much “income” is earned every day in the stock market from people selling shares.”
So just to clarify, is the profit generated from, say, day trading considered income?
AP, I don’t know how the GDP accounts reckon it, but in terms of economic theory yes, of course profit generated from stock speculation is income. It’s a capital gain.
If I buy a share at $100 and sell it at $110, my income during that period is $10. The point I was making in the post was that my income is NOT the full $110, which is what I think MF wants to say.
There I two things I don’t understand and I’d appreciate if someone could help me out here.
First, all this talk about arithmetics and equations seems postulate that “the economy” is an isolated system. But what about all the money owned or lend to foreigners. Where does that come into the equation. And what if – in Dr. Murphy’s story – Cathy was a Chinese and the American economy consisted of Larry and Willy. It seems to me that in this case it is possible for Larry’s and Willy’s economy to have a net reduction in debt if Cathy’s economy had a net reduction in savings. I know that this argument is not Dr. Murphy’s point, and maybe I am just missing something, but even if Krugsman’s arithmetics are correct, I don’t see why Americans cannot deleverage and let others (Chinese and Japanese) dissave. So any explanation would be appreciated.
Second, suppose it was true that net debt reduction always reduces income, wouldn’t it follow that there is no other way to generate net income but for someone else to go into debt? So couldn’t you refute the argument by showing that there are economies that do not have debt but still have income? Are there such economies?
Again, maybe I am just not understanding the basics of economics.