Wednesday, January 29, 2014

Chicken and Egg

--Updated 2/10/14--I don' think this post is 100% correct or even 75% correct.  I think excess reserves are a function of the fed funds rate hitting 0% rather than QE.  More on this later...

"The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design." - Friedrich Hayek

People forget that we went off the gold standard in 1971 not due to furious public or academic discourse weighing the pros and cons of fiat currency, but out of panic.  The dollar was in trouble and it seemed everybody around the world wanted to exchange their dollars for gold.  Even France, whom America felt they rescued just 30 years ago from big bad Hitler wanted out of dollars.  Thus, if everybody wanted gold for dollars, the most obvious thing to do was to no longer allow it and that is exactly what Nixon did.  But Nixon and his advisers never meant the so-called Nixon Shock to change the fundamentals of our monetary paradigm--he just wanted dirty foreigner hands' off our precious gold.

However, there was a fundamental change: with money supply tied to gold, if you wanted to increase the amount of money, better find more gold, lest you officially revalue your money to gold every so often.  However, this required precise forecasting--a little bit too high or low and you end up with crippling inflation/deflation.  This was the shortcoming of gold standard: the lack of quick flexibility to respond to volatile conditions.  Fiat currency unintentionally solved that problem for them and that was Hayek's insight.  All Nixon wanted to do was to count his gold every night and he accidentally created a fiat paradigm that even today economists cannot agree on just how it exactly works.  This is not because it is hopelessly complex, but because it was unplanned.  If one does not expect a change, even when a change happens, he may not see it.

One of the most glaring examples is this simple question: how is money created in the private sector?  In a sense, it seems impossible for money to be created or destroyed as one person's income has to be another's expense.  The answer is obviously that bank debt is freely used as money.  If I have $100 in a bank and the bank lends out $100 to Bob, I still count my $100 and Bob has another $100 to spend.  Sure he eventually has to pay it back, but in a healthy economy, loans are created every day, creating the growth in money supply.  In a lark, I did some charts:

FRED Graph

Green line is m2 money supply.  Blue is all bank loans and red is all bank deposits.  So what happened in 2008?  Why the big gap between loans and deposits?  How is this even possible?  Where are people even getting deposits without corresponding loans?  Sober Look astutely pointed out that the gap reminded him of excess reserves and adding back the excess reserves, the lines are harmonious once again.

FRED Graph

But what are excess reserves?  And why did they not even exist until 2008, exactly when QE started.

FRED Graph

So I compared it to the Treasuries held by the Fed:

FRED Graph

Pretty good.  Red is treasuries held by the Fed, blue is excess reserves.  This simple chart supports the notion that excess reserves are just an accounting identity due to QE.  Those who say excess reserves prove banks aren't lending are wrong.  Excess reserves seem to exist because the Fed paid depository institutions reserves and took their Treasuries.  These reserves thus represent the part of their balance sheet that was supposed to be risk-free income from Treasury coupons.  In fact, you can make the argument that QE is deflationary since the private sector is losing out on these interest payments.  Regardless, the relationship between loans and deposits and money supply still stands the test of time and offers a clue into the ultimate chicken and egg problem: what comes first, the loan or the deposit?

Superficially, it seems obvious that a bank needs to take in deposits first in order to lend that out.  We have terms like money multiplier and reserve ratio to describe this effect: deposit $100 and see how many times this $100 gets lent out, ie multiplied and what percentage of that $100 needs to be kept as reserves.  However, these seem to be meaningless limits.  Banks will always lend, provided they believe the applicant will be able to pay the money back and their capital structure can support it.  They will never have trouble looking for reserves to satisfy the reserve requirement.  In normal times, they can go to the interbank markets and pay the fed funds rate to borrow from other banks.  Because the Fed has decided to target the fed funds rate, it will always make sure there is the correct amount of reserves in the system to correlate to that targeted interest rate.   Even when the economy goes to shit and loans start going bad/people draw large amounts of cash from their deposit accounts, reserves are easily made up by going to the Fed's discount window.  This is in addition to the Fed typically lowering rates during bad times anyways, which incidentally raises reserves in the private sector (as the Fed pays the private sector reserves to buy up short term bonds).

This is where excess reserves come in.  The Fed realized they had to pay interest on these reserves that are parked at the Fed as any excess reserves mean the Fed Funds rate is irrelevant.  Suppose Bank A sold some Treasuries to the Fed and thus has some excess reserve.  It tries to lend it to other banks, but no banks has use for it because creating loans also creates the requisite reserves in the system as a whole.  Thus these excess reserves due to QE has no demand and would drive the interbank lending rates (fed funds rate) to effectively zero.  This is why the Fed had to pay .25% interest on excess reserves: it has become the de facto fed funds rate.  Without the .25% and with the existence of excess reserves, fed funds rate will always be 0%.

So why is this all important?  If you understand that the most important variable of business cycles is debt, then money supply becomes one of your most effective tools to brunt the deflationary effects of debt in the downturn.  Recessions happen because households and corporations repair their balance sheet via dutifully paying off debt rather than borrowing and spending.  The fiscal and monetary powers-that-be can in the short term alleviate this process by inflating and making debt easier to pay back.  And we all know the easiest way to inflate is to increase the money supply.  Thus it is crucial to understand just how money supply goes up or down.

In essence, it seems the market for money, like all markets, is an intersection of supply of demand.  However, once the Fed sets an interest rate, there is essentially an unlimited amount of supply of money offered to the private sector at that interest rate until the Fed decides to set a different interest rate.  In 2008 when deflation was threatening to turn to Depression, the Fed essentially lowered the price of borrowing to almost zero.  Economics 101 tells us demand to borrow should have skyrocketed and in turn cushioned the drop in money supply from all the debt going bad.  However, the private sector's balance sheet, particularly the households, were in such disarray that they could not even afford to borrow at historically low rates.  Thus money supply remained lower than it "should have" as unlimited supply at low prices mean nothing if there is no demand.  This in turn led to the flirtation with constant deflation.

Taken one step further, this explanation means any supply side monetary stimulus, such as increasing monetary base, is doomed to fail in these types of situation.  It was not the case that there was not enough credit to go around.  It was the case that nobody wanted the credit.  To be fair, Bernanke did all he could.  He even jawboned Congress to do more because he understood the viability of fiscal policy to directly put money into households' bank accounts, which in turn can be used to pay off debt over his own flaccid monetary policy.  Keynesians were right in 2008 to push government intervention in the shape of bigger deficits, provided their goal was short term growth.  Monetarists are also right that going forward now, monetary policy will be more helpful as balance sheets have been largely repaired.  Meanwhile, Austrians do their best Grumpy Cat impression to chide both camps that their incessant meddling was what got us into this mess in the first time.  Only the neoclassical economist continues his research, blissfully unaware that the foundations of his worldview are being eroded daily under his feet.