On Liquidity Holes & Retail Liquidation
The Great Rebalancing
Consider that crypto has only existed in a 13-year window where central banks conducted a $20trn balance sheet expansion.
We now expect the rate of change of liquidity to go negative.
What are the cause / effect relationships of QT and what’s the path?
Central Bank Balance Sheets
Policy support moves to quantitative easing (QE) when interest rates reach 0%.
QE occurs when central banks buy assets - typically government bonds - and deliver money into the economy.
Since the seller of a bond is unlikely to buy bonds again - that money trickles into cash or riskier assets.
Because cash has yielded 0% for much of the last 13 years we can conclude QE money has trickled into riskier assets like corporate bonds, equities and crypto.
QE has the implicit goal of delivering money into assets further along the risk curve which increases the valuation of financial assets.
Financial assets are claims on a country’s economy
A stock is a claim on a company's cash flows with within an economy
A government bond is a claim on the country’s tax receipts
Real estate is a claim on the expected best use of land within a country
QE increases financial asset values vs the economy.
Some reflect this relationship using stock market capitalization / GDP which has reached multi decade highs until the 2022 market rout.
QE can be simplified as an attempt to increase the financial claims on the economy.
Quantiative tightening (QT) has the goal of lowering / liquidating financial asset claims on the economy.
QE/QT Interest Rate Equivocation
Equivocating QE policy support into interest rate terms is intuitive.
If we do $100bn of QE - what is the equivalent interest rate change that creates the same liquidity / financial conditions impact?
Wu-Xia show that during 2008-2019 QE was equivalent to -3% policy rate — they call this the “shadow rate”
This means the QE policies were equivalent to a -3% interest rate or getting paid to borrow money.
This gives us a quantitative framework for what type of rate of change QE provided markets and how withdrawing it will tighten markets.
Withdrawing liquidity is path dependent on growth and inflation.
Concluding QT will occur until a target is met undermines historical policy reflexivity.
If rising interest rates and QT create a sufficient slow down in growth to lower inflation then we've touched an equilibrium.
For now — the Fed plans to withdraw $95bn per month or nearly twice the pace of the 2017-2019 QT cycle.
This tightening is simlar to the level of accodomdation given during QE3 in Fall-2012.
The current QT path implies a $1.14trn annualized withdrawal.
What is the duration of this program likely to be?
Pre-covid the Fed's balance sheet was $3.5trn - now its $9trn
One could argue we should withdraw $5.5trn of liquidity which implies the QT program should last 5 years.
But that ignores nominal GDP growth and the inelasticity of the Fed's balance sheet to inflation.
For instance, it’s unclear why the Fed can’t stop at $6.5trn since QT is motivated not by academic purity but its influence on growth / inflation.
$7trn seems like a reasonable target but its merely a guess.
A $9trn to $7trn balance sheet reduction implies a 2 year QT program at the $95bn/month guided run rate.
Sources and Uses
QT implies the market needs to absorb bonds from the Fed.
The market also needs to absorb primary treasury issuance from the deficit.
Buyers of treasury bonds include:
Foreign holdings of treasuries have been flat for over a year.
Given that hedged yields of Euro/Yen are low — it's possible to rule out foreigners being large buyers of treasuries.
JPY / Euro Yields Hedged Back into USD
Therefore issuance / Fed secondary sales must be absorbed by domestic sources.
Banks have been large buyers of bonds / mortgage backed securities because they offered a yield pick up to the 0% short terms rates.
During the last QT cycle from 2017-18 we saw a significant slow down in bank ownership of treasuries.
If this holds again — that means the bonds must be absorbed by non-banking sources which is mainly households.
Constructing a sources / uses matrix we see that households will need to absorb $4.2trn in liquidity assuming a QT program of 2 years at $95bn/month.
Federal deficits are sourced from the Q2 TBAC prsentation.
These numbers can change based on duration / size of monthly QT but the point is the bond supply / demand liquidity hole is likely to come from households.
To be clear — very similar math existed in 2018 before the year-end crash.
The Fed was able to pause QT and economic / market conditions normalized as soon as the Fed paused.
The key difference today vs 2018 is that in 2018 there was really no reason to raise rates / QT but for 'having bullets' for the next downturn — there was no binding constraint to keep policy tight.
Inflation is the binding constraint today and therefore all path scenarios are contingent on realized inflation its perceived rate of change.
Household Balance Sheets
QT implies tight liquidity conditions using the shadow rate model.
Even more - the supply/demand imbalance will have to be met by US household.
We've sized a $4.3trn liquidity need from households.
This will come from
Households still hold excess cash above trend from the Fed’s QE policy.
Here we can eye-ball that $4trn of liquidity withdrawal get’s us back to trend.
In addition households can liquidate financial assets.
Here we can see household directly held equity portfolios grew $11.4trn from YE 2019 to 2021 — after the recent rout the gain is $5trn.
It’s likely households will raise liquidity via cash and equity sales to rotate into the treasury liquidity gap.
Crypto is an obvious source as well.
Price Elasticity of Flows
If households are to balance the liquidity gap with sales of financial assets its important to examine that impact.
If there is a $1 flow into a equities - what is the impact on price?
Several methodologies in recent years suggest that answer is at least 1 and more likely 2-5x
We’ve established the following
QT’s intent is to lower values of financial claims on the economy
QE at its height in 2009-19 cycle provided a -3% interest rate benefit — QT will provide the inverse
Sizing QT duration and scale implies ~$4trn of liquidity withdrawal
Foreigners and banks are unlikely to be sources of funds to meet treasury demand
Households will need to liquidate cash / financial assets to meet treasury demand
Studies of flows into equities suggest every $1 represents a 2-5x impact on price