Inflation and Asset Price Responses, and the (Non)Correlation to the Central Bank Policy Stance
I was asked recently to reconcile my position that Quantitative Easing (or similar activity) is not inflationary. Especially when viewed against a ten-year chart of virtually any commodity denominated in US Dollars.
This is a very loaded question and touches on all aspects of economics and the markets (and yes, these are very different things). I also think the framework for the question and its comparison is also incorrect because it make the implicit assumption that the main driver of this price rise is US Monetary Policy (i.e. the US Federal Reserve). This is also very interesting because I have been taking a lot of flak for my positions (both on CAPS and MTaA).
If I was asked this same question two years ago, I would likely have given a very different (and likely far more rigid) answer.
First let me say what I am not:
I am not a Keynesian
I am not an Austrian
I am not an MMTer
I am not a Fed apologist
I am not a Gold Bug
I am not a Paper Bug
I am not a Permabull
I am not a Permabear
I am not dogmatic on any economic theory or economic analysis
So who am I and what do I believe?
I am a student of the market
I am an avid amateur who studies all economic schools of thought and picks out that which makes sense
I do believe macroeconomic analysis is a useful endeavor
I do believe technical analysis is a useful endeavor
I do believe cycles analysis is a useful endeavor
As such I will never take any argument that someone makes completely on faith or at face value. By the same token I will not dismiss any analysis out of hand without considering its merits. This also means that as I incorporate new ideas into my own thinking process, I will let go of previous misconceptions if I believe that they are in error. And I do think some of my previous macroeconomic analysis has been in error. As always, I fully reserve the right to be wrong, and to change my stance upon the discovery of better (more accurate) information and ideas.
Like I have said before (Regarding Economic Debates and Opinions: The Fallacy of "Purely Objective" Analysis) the study of economics is a study in subjectivity. As such there are only shades of gray and there are two (or more) sides to every position. Anyone who says otherwise is trying to sell you something.
And I have nothing to sell.
I am not trying to convince anybody of anything. I have an analytical mind and I lay out my analysis for anybody to consider. My only goal is to point out things that may not be obvious and to further reasonable discussion and observations regarding macroeconomics and the markets. Maybe you agree with me. Maybe you think I am full of crap. Maybe you don't care either way. I am fine with any of these reactions. As long as one of my posts caused you to think, or to perhaps look at something in a slightly different way, then it did its job.
The reason I say all of this preamble is because I will likely discuss things in a very unpopular light. I might even have some parts of my argument where you are nodding your head in agreement and then I will get a "WTF!" in the very next paragraph. There are a few mainstream ways to looking at inflation vs. deflation, commodities vs. stocks, deficits vs. surpluses, stimulus vs. austerity, and I am convinced that most of the mainstream thoughts are partly (if not completely) incorrect.
You might expect me to throw down some gauntlet, but that is not my intention. Like I said above, I am not trying to convince anybody of anything. And at the risk of sounding harsh, I really don't care if you agree with me. Garnering agreement is not the point of this point. I am just laying down thoughts for consideration, that is the only point of this post.
Understanding the Fiat Currency Monetary System and How the Banking System Works
This is the most important aspect to understanding macroeconomic consequences and to attributing proper causes and effects (and to disregarding fallacious 'causes' and 'effects'). I think the vast majority of macroeconomic analysis is incorrect because it either doesn't properly understand the monetary system, or it assumes the monetary system works now the way it did under a convertible currency system.
Let me first say that I am not agreeing that this is the best monetary system to have. My interest in studying it is to recognize how the system behaves. The point of this post and analysis is not to explore what 'should be'. It is neither praise nor condemnation of our current system. But denying the reality of our system would make me both a bad analyst and a bad investor. And I have no desire to be either.
There are some posts that I have written which are very good background for this topic, and I highly suggest reading those first. I will be doing a lot of summarization in this post (it will be quite long enough as it is).
(1) The Matter of Deficits, Sovereign Default, and Modern Monetary Theory, explores the history of our monetary system, from the gold standard to fiat money, and describes the role of government bonds in both systems.
(2) Follow Up: QE is not Inflationary, Thoughts on Risk Asset Instability, describes why banks are not reserve constrained, and shows how OMO (of which QE is nothing but a variant) actually works and why it is nothing but an asset swap.
(3) One of the smartest comments I have read yet regarding Quantitative Easing, Bank reserves are not lent out to the private sector. QE adds banks reserves in exchange for bonds (the swap) and does not give the banking system any more ability to do anything that it couldn't do before.
The first thing to understand is that the money multiplier model is a myth (see here and here). There are so many that point to the massive increase in bank reserves (as a result of QE) and say that 'this is increasing the monetary base and will inevitably be inflationary'. This displays an extreme lack of understanding of how our banking system actually works. If you do want to know how the banking system works (and to understand that banks are not reserve constrained but are rather capital constrained, and targeting reserve levels does nothing to increase the propensity to lend) please read references (1) and (2) above.
What policies like QE and interest rate targeting assume is that monetary policy can drive behavior in the banking system (i.e. that the behavior is driven by the supply side of the equation). And we have seen time and again that this assumption is invalid, that the primary driver of activity in the banking system is from the demand side.
Were low interest rates the primary cause of the housing bubble? Or were lax lending standards? I believe that it is incontrovertible that it was the latter. If someone has a $30,000 income and can buy a $400,000 mortgage (based on no down payment, and either dubious paperwork or loan officers that are very aggressive) on the belief that they can flip the house for a quick profit, would it have mattered if the interest rate was at 8%? Conversely, if interest rates were at 0%, but a 20% down payment was required would people be buying able to buy houses to flip as we saw during the bubble?
My point of discussing all of this in the context of this section of the post is to disabuse you of the notion that Monetary Policy is 'all-powerful' or that it is the primary driver of what happens in the economy. It is a popular theory, and I also believe that it is incorrect.
What this means is that the Federal Reserve is not the all-important entity controlling the markets. Contrary to a lot of stated rhetoric, I think the Federal Reserve is largely an impotent organization. They can 'grease the skids' when inflation expectations are high and they can try to 'jawbone' higher inflation expectations, but at the end of the day their toolset (OMO, QE, etc.) targets only the supply side of the banking system and can purchase assets that only already exist.
I recently had a comment leveled at me that 'I just don't get it', that this is not the way this works, that 'Ben could drop money from helicopters like he said in his speech'. ... Except that he can't. The Federal Reserve cannot arbitrarily print money and drop it on the markets, either literally via helicopters or though the banking system.
Suggesting so displays a lack of understanding of how the banking system works. Everything that happens in the banking system is completely horizontal in nature. Every time the bank issues an asset (a new loan or creation of 'money from nothing') it must also record that loan as a liability. This means that all transactions within the banking system net to zero, and hence the description 'horizontal'. This includes the Federal Reserve. It can change the composition of assets within the banking system (it can create reserves out of thin air in exchange for Treasuries) but it doesn't change the net amount of assets within the banking system.
The only party in the banking system that can can the net amount of assets is the US Treasury.
It is the only entity that can create money 'vertically' (that is, create an asset in the banking system with no corresponding liability). This is critical to understand in order to view the entire monetary system properly. So, if this doesn't make sense to you yet, read ref (2) above and here (read the whole thing, but the section titled 'Mechanics of Federal Spending' is extremely relevant). Conversely, the US Treasury is the only entity than can drain money from the banking system. When the US Treasury spends more than it taxes, money (specifically Federal Reserve Notes = fiat currency) is created. When the US Treasury taxes more than it spends, money is destroyed.
When you realize that full scope of this concept - that all transactions in the banking system net to zero and that the US Treasury is the only entity can add net assets to the banking system - you immediately come to the correct conclusion that Federal Government spending is the source of all money in our currency system.
In order to hammer this point home, consider this question: "What would happen if the US Federal Government issued a one-time 100% asset tax?". Everything would be liquidated and owned by the government. So after this event, how would one acquire a US Dollar if all assets were owned by the government? The government would have to spend new dollars into existence so that the private sector could accumulate them. I am stating this merely to illustrate the point that the US Government is sovereign issuer of the US Dollar and as such is the source of all US Dollars. I am not saying this is good, I am not saying this is bad, I am saying that in our current monetary system this is a fact that should be self-evident.
So let's return to the 'money helicopter drop' idea.
Like I have said repeatedly, the Federal Reserve cannot arbitrarily print money (i.e. cannot increase net financial assets in the banking system, it can only create reserves [out of thin air] in exchange for assets that already exist). But there is one entity that can: The US Treasury.
The US Treasury acts under direct control from Congress. This means that a 'helicopter drop' is NOT a monetary operation. It is a fiscal operation. The US government could instruct the Treasury to credit everyone's bank account by $10,000. Or if it wanted to put on a big show, it could actually print the money and 'drop it from helicopters'. But think about these mechanics. The Federal Reserve actually prints the money. So what happens is that the US Treasury instructs the Fed to increase its account size at the Fed and to print money equal to that account size and transfer it to the Treasury (and would then give it to the Air Force with instructions on where to drop it). So again, the Federal Reserve cannot arbitrarily print money. It can only change the composition of assets that already exist. And in this case the Treasury is the only entity that can arbitrarily increase net assets.
This might seem like I am being facetious, but I am not. This leads to a very important point:
Where does inflation come from? (Hint: It's not 'Debt Monetization')
It comes from Congressional spending.
Monetary inflation and price inflation are not the same thing. They are very different phenomena. One does not always beget the other (especially in terms of relatively short term timing), I will discussing this in more detail shortly. But the very long trend throughout US history of deficit spending goes hand in hand with the long term trend of inflation.
At this point, I will not call inflation 'bad' or 'good'. The point of the post is to simply understand the current system how how things like inflation manifest themselves. Because you can't have an intelligent discussion about outcomes and risks if you misidentify/misunderstand causes and sources.
It doesn't come from the Federal Reserve. I know it is extremely popular to thing of the Fed as some evil organization of banking dictators that can arbitrarily tip the dollar into hyperinflation due to reckless 'money printing'. But that view is completely incorrect.
So why is it so popular to think that? Why do analysts inevitably link inflation coming from the Fed and Debt Monetization?
A quick history lesson is in order. I describe this in much more detail in ref (1) above.
Under the Gold Standard the US Government was required to have all currency issued backed by gold. When the government need to finance spending (say to pay for a stimulus or a war) in excess of held gold reserves, the US government would issue bonds. In this case, the term financing is accurate, US Government bonds under the gold standard were most definitely debt. The US Dollar, being convertible into gold, had that constraint on it. Which meant that the US government was most definitely revenue constrained. That taxes were a repayment of Dollars such that the outstanding liability of conversion back into gold was reduced. Under Congressional mandate, all spending in excess of taxation must be accompanied by US Government bond sales. Sometimes the non-government sector would buy the Government bonds. But sometimes they would be bought by the Federal Reserve. And this is where the term 'debt monetization' comes from. Under the gold standard, US Dollars were convertible into gold which made US Government bonds, which were convertible into Dollars, indirectly convertible into gold. So when the Federal Reserve would issue currency in exchange for those bonds, the debt would become 'monetized'.
That is very different that what happens today in a fiat currency system, and why 'debt monetization' is (at best) an inapplicable and (more often at worst) highly misleading concept.
Since 1971 the gold window was closed. And as I showed in ref (1) this was accompanied by some very ad hoc activities. Namely the issuance of US Government bonds. Since the US Dollar is no longer convertible into gold, it is no longer debt in the literal sense that it had under the gold standard. Since 1971, the US Dollar is one thing only: a tax credit. I have seen many people try to conflate the idea between debt and a tax credit, but this again shows a lack of understanding of how significantly things are different under the two monetary systems. This means the US Dollar has no intrinsic worth. The Federal government issues currency and sets taxes payable only in that currency. It (via the Treasury) is the only source of said currency. It's value is literally fiat. This means that operationally the US Government is never revenue constrained (i.e. that it will always be able to credit accounts without any liabilities via the Treasury to purchase assets and to service interest payments denominated in US Dollars), and it never needs to 'finance' it's spending. This means that under the current monetary system Federal taxes 'fund' nothing and the US Government bond issuance 'fund' nothing.
So then why does the US Government issue bonds now when it deficit spends? Because it is a holdover from the the gold standard days. Nixon closed the gold window but did not think through all of the implications of transitioning to a fiat monetary system. And so original rule issued under congressional mandate was never changed by a new congressional mandate and hence we have this anachronism still in place causing all sorts of confusion with people trying to understand the monetary system.
So then what do US Government bonds 'do'? They allow the Federal Reserve to maintain its target short term rate. Deposits within the banking system create reserves. A bank manager has a few options with what to do with those reserves, after using a portion of those reserves to ensure balances are cleared and other maintenance activities, and to maintain reserve requirements (and it is worth noting that many banking systems in other countries have no reserve requirements at all). They could let it sit in the vault as cash (earning no interest). They could keep it on account at the Federal Reserve (which relatively recently now earn a mere 25 bp, they earned 0 for a very long time before that). They could lend the reserves overnight (at an interest premium) on the interbank lending network to banks that require reserves to meet requirements. Or one of the few options beyond that is the purchase of US Government bonds. Reserves are not lent out. With the purchase of the bonds the banks are able to earn more interest in a very liquid asset (the US Treasury market is by far the biggest on the planet).
This is how this manifests in the current system: The Federal Reserve has a mandate to maintain short term interest rates consistent with its policy goals. If the banking system has a system-wide deficit of reserves relative to requirements, then competition overnight on the interbank network will drive up the short term interest rate. The Fed reacts to this by buying US Treasury bonds on the open market and prints reserves (out of thin air) in exchange for those bonds. The result is a system wide increase in reserves and a system wide decrease in the amount of US Government bonds (an asset swap). This allows the Fed to put a cap on the short term interest rate relative to its policy goals. On the flip side, If the banking system has an excess amount of reserves relative to requirements, then competition overnight on the interbank network will drive down the short term interest rate. The Fed reacts to this by selling US Treasury bonds from its portfolio on the open market. The banks trade their reserves in exchange for the bonds. The result is a system wide decrease in reserves and a system wide increase in the amount of US Government bonds (again, nothing but an asset swap). Since the Fed is the entity that 'printed' those reserves to begin with, they disappear 'into the ether' when they are returned to the Fed's balance sheet.
The only reason for a US Government bond to exist in today's banking system is to allow the Fed to meet its short term lending rate based on OMO. In fact, there is no real physical reason for this song and dance. The same policy objective could be accomplished by the Fed setting a remuneration rate at their policy interest rate. The only reason in the current system that banks juggle between reserves and Treasuries is the opportunity cost in obtaining a higher yield with the excess reserves. If the interest rate and the remuneration rate were the same, then there is no opportunity cost differential and the banks would find themselves at no disadvantage to holding on to excess reserves. In this fashion the US Treasury could stop selling 'debt' and the system would be significantly cleaner, and much less confusing.
Once you view reserves and US Government bonds in the proper context as outlined above, you come to the realization that in the current monetary system, bonds are nothing more than a savings account at the Federal Reserve. That's all they are. The US Government issues bonds. The US Government is not revenue constrained and can always 'afford' to service interest payments. The Government instructs the Treasury to pay the interest. The Treasury instructs the Fed to 'print the dollars to service the interest' (i.e. add new numbers to a spreadsheet) via crediting a bank account at the Fed. That is the process. All of the manipulation of 'servicing the debt' happens on the Fed's balance sheet. When an entity sells a Treasury bond the Fed takes the Treasury bond which is logged on the US Treasuries account at the Federal Reserve, removes it from the US Treasuries balance sheet and credits that entities bank account. It is all simply movement of numbers between accounts at the Fed, in exactly the same fashion that one moves money between a checking and savings account at a bank.
Why the Federal Reserve Follows Both Inflationary Cycles and Credit Cycles and Doesn't Lead Them
As I explained above, the Federal Reserve manages the level of reserves in the banking system consistent with it's policy interest rate goals, and can do nothing to affect the level of net financial assets in the banking system.
Moreover, as also discussed above, the Federal Reserve via Monetary Policy is a blunt instrument targeting only the supply side of lending equation. It can force a surfeit of reserves (as it has recently) but there is absolutely no correlation between the level of excess reserves and horizontal lending activity in the banking system.
Under neo-classical economic theory, a large amount of excess bank reserves will lead to an increase in bank lending via the Money Multiplier Model. The very abbreviated version goes like: In a Fractional Reserve Banking System the total amount of loans that commercial banks are allowed to extend (the commercial bank money that they can legally create) is a multiple of reserves. Which means the higher amount of reserves, the more loans / credit money can be created.
However there are a number of issues with the Money Multiplier Model in terms of being the driver of forcing monetary policy. It basically is a 'supply side' theory in which the bottleneck is the banking system. By increasing the reserves in the banking system, more loans will be made because there is a higher capacity to create loans. This does not hold up with empirical observations (see here).
Operationally, banks are never really reserve constrained. Reserves only need to be met against a 14 day average, which means if you are a bank you can make a loan absent the required reserves at any point. Reserves can be acquired either through interbank lending or the discount window. Moreover, reserves need to be carried against demand deposits (checking accounts). Sweeps have been routinely implemented to move demand deposits into a special purpose savings account (a savings account is not a demand deposit) obviating the need for higher reserves.
In a fantastic paper by Austrian Economist Vijay Boyapati (see here) observation has shown that bank reserve creation by the Fed actually lags banking sector lending
Given that commercial banks are effectively operating without a reserve requirement and that loan issuance is not constrained by reserves, it would be sensible to reconsider the temporal causality posited by the money multiplier theory of lending. If the causality were correct, one would expect changes in reserves to precede changes in the issuance of credit, ceteris paribus. Citing an empirical study on business cycle statistics conducted by the Federal Reserve, Steve Keen explains that the opposite is true:
…rather than [reserves] being created first and credit money following with a lag, the sequence was reversed: credit money was created first, and [reserves were] then created about a year later.
From an Austrian perspective, an empirical argument based on a temporal correlation is not definitive proof of an underlying causality — although it may be illustrative and suggestive of that causality. An explanation for the counterintuitive temporal sequence is provided in a Federal Reserve study of the institutional structure of the US banking system since 1990, conducted by Carpenter and Demiralp. They demonstrate that “reservable liabilities fund only a small fraction of bank lending and the evidence suggests that they are not the marginal source of funding, either.” Their point is that when a bank makes a loan, the matching liability used to fund the loan does not need to be reserves created by the Fed.
These two important observations [1) Reserve creation in a 'normal' credit cycle follows, not leads, created loans, 2) Banks are never operationally constrained by reserve levels when making a loan decision, reserves are always found after the fact] should make you question and doubt anyone making the claim that the large increase in excess reserves will result in inflation.
Why the Federal Reserve is a Redundant and Unnecessary Institution and Should be Abolished
I just spent several paragraphs explaining why the Federal Reserve is largely an impotent institution from the perspective of forcing any real macroeconomic activity. So what's with the abolishment rhetoric?
This is one of the few times in this post that I will suggest a course of action, and not simply illuminate the current workings of the system, because the Fed's presence, while impotent, is not innocuous.
Please read reference 4 below for more details:
(4) The US Treasury and the Federal Reserve: Redundant Institutions
Assuming that we will have a fiat currency system for some time (and there is no reason to think that we won't) then having this formally redundant split between the Treasury and Federal Reserve obfuscates the fact that they are really two halves of the same balance sheet. There would be significantly more transparency in how our system works if we did away with this anachronism.
Moreover, the possibility (and likelihood) for corruption is high during times of crisis. The Fed took all kinds of toxic mortgage garbage on its balance sheet (yes, in exchange for reserves printed out of thin air) to save the banks. However, the Fed didn't buy everything nor did it buy unilaterally. It bought selectively and potentially with bias in mind. It has a very cushy relationship with several banks.
So as explained in the monetary operations section above, there is no advantage to be gained by pretending the Fed and Treasury are completely independent institutions when that is obviously not the case, and there are many potential downsides. The upside is a massive gain in transparency, and I think that should be very welcome.
Business Cycles / The problem with attributing a specific period of price rise to 'inflation'
Contrary to all of the statements from mainstream economists in the 1990's, the business cycle is not dead. It was not 'conquered'. The 'Great Moderation' is a myth. The reason why these pronouncements were made was because inflation was down-trending and private sector productivity was increasing (although in the latter half of the 90's, this was not true because of the increasing financialization of the US economy, which is inherently unproductive). This was all attributed to the 'correct balance of monetary policy'.
But as I discussed above, Monetary policy is only useful at greasing the skids during times of moderate to high inflation expectations. It is completely ineffective during a balance sheet recession. And the current balance sheet was not at all predicted by mainstream economists, especially not those in charge of US Monetary Policy.
Because all of them thought (and continue to think) that the economy can be managed from the supply side. And I think that is a very bad assumption.
For every seller there must be a buyer, and for a rational economy (and by extension a rational market) to exist all market actors must behave rationally and respond 'correctly' to all available information, making the economy and the markets 'efficient'. This is obviously not the case (if it was ever the case at any point in time).
Market actors are not always rational, and hence economic behavior cannot be fully determined by supply side actions, because the demand side is not always rational (which further compounds the lie because it assumes that all actions taken by the supply side actors [a.k.a the government economists and the Fed] are inherently rational, which I think is a crock). Let us return to the housing bubble example. Historical data showed the even in the middle of the bubble (before things got really crazy in 2006-2007) that never in America's history did the ratio between home prices and incomes become so large. There were numerous statistics that anyone with an internet connection could have browsed that gave a hint at the unsustainability of the current trend, and the information clearly pointed against a trend continuation. So if all of the actors on the demand side of the equation were rational, how could a bubble form?
The urge to speculate is a powerful phenomena and we see evidence of this over and over again.
If markets were completely efficient, why would there be any P/E variation in the market? Maybe we could explain a minor variation in P/E via cyclical adjustments in the market, but that is not what we see. A 100 year chart of Price/Earnings for the S&P Composite / S&P 500 shows that P/E changes from 5 to 40 on a cyclically adjusted basis. This is not rational.
Markets are cyclical because social mood is an inherently cyclical phenomenon.
Consider this from a slightly different angle: If the long term trend of inflation is caused by the long deficit spending position of the US Government (and I believe it is), why would there be any cyclical variation in prices? Why wouldn't prices march in lock step with the size of the accumulated deficit?
Because markets and market participation is cyclical.
This gets us to the answer to the original question.
- Is inflation directly attributable to the position of the Federal Reserve? NO.
- Is inflation directly attributable to the position of the fiscal policy stance of the US government? LONG TERM YES, SHORT TERM NOT NECESSARILY.
- Are markets cyclical? YES.
- Is cyclical participation combined with long term inflation trends a more accurate description of market behavior? YES.
This is why I have a huge problem when anyone tries to attribute the rise in commodity prices over any period of time to 'monetary policy/money printing/quantitative easing/etc'. It doesn't jive at all with how our monetary system works. It doesn't jive at all with how actors participate in the markets. The monetary stance of the central bank controls the supply side of the lending equation. That's it. It does not either implicitly or explicitly increase the propensity of borrowers to change their position relative to their own balance sheets.
Cycles in any asset class (commodities, corporate investment, housing, etc.) happen independently of the central bank policy stance. They are much more susceptible to changes in the fiscal and regulatory environments (i.e. Congressional action or inaction). Attributing credit booms and busts primarily to central bank activity displays a misunderstanding of both the banking system and the role of the central bank within the monetary system.