Monetarist normies will tell you government spending & money supply, but they are wrong! Only #MMT understands this, and the data proves it.
So let’s do a little data science to see what the real cause of inflation is!
A thread ...🧵 1/10
Our first step to determine what drives inflation is to plot our candidate inflation variables against CPI (see 1/10) and take a look at the time series. It won’t require a PhD in stats to see there is a correlation between each variable and CPI. So let’s dive deeper. 2/10
To do so, we’re going to run two machine learning analyses, Mutual Information (MI) and Granger Causality. MI will measure the amount of information contained between our variables and CPI, Granger Cause will tell us if our variable time series ‘causes’ the change in CPI.
3/10
First step, we’ll determine MI using @scikit_learn mutual_info_regression:
What does this tell us? Relative to the entire list, Fed Funds contains the greatest amount of information about CPI, but each of our variables tell us far more than our baseline, local temps.
4/10
In other words, these are all reasonably linked with inflation and may be causal drivers of CPI. And that’s where Granger Causality steps in!
5/10
To run Granger Cause we first need to make our data stationary, meaning we need to have the mean and std dev constant and remove seasonality, so now our data now looks like this:
6/10
Now its time to run the Granger Cause test. We’ll use @statsmodels grangercausalitytests which will tell us if our variables are useful for forecasting CPI. And what do we find 🥁:
7/10
10-year yields & Oil Inventory Granger Cause Changes in CPI
Govt Spending, M2, & Fed Funds Do Not Granger Cause Changes in CPI
CPI Granger Causes Change in Deposits
8/10
The results seem to tell the same story that @wbmosler & MMT have long argued. Inflation is a not a function of government spending or money supply increases. But its rates and real output that tell much more of the story than what is currently understood by the mainstream.
9/10
Now that we’ve done the hard work of analyzing the data we can confidently say that only #MMT understands what drives inflation.
Follow me and my partner in crime @MathWillSuffice Check out our YT channel and if you're an active investor check out our patreon (🔗 in bio)
10/10
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Want the best Bull case for '23 one that only #MMT understands: A massive wave of billions in free $ is set to hit the private sector via interest payments from the Govt. Before we understand why this will be a + for stocks we need to understand what caused the '22 selloff 1/6🧵
What caused the selloff in 22? A collapse in government deficits. MMT gets that govt spending is private sector saving. Once govt spending reversed in late ’21 the priv sect was reliant on an already flimsy credit cycle 2/6
A credit cycle that ultimately couldn’t hold the weight of the disappearing deficit flows. But this trend is reversing in a big way! With the Fed driving rates higher, we’ll likely see net monthly interest payments nearing $100B per month in 2023. 3/6
Loans Create Deposits – This is a core tenet of #MMT and a grossly misunderstood dynamic in economics and finance – Here is a simple balance sheet operation to understand how banks create money and a brief introduction to the effects of this process 1/5
To set up our example, lets assume I want to buy your house for $250k. To do this, I need a mortgage, which the bank creates, and in doing so will also create a new deposit for you on their balance sheet, this mortgage is both an asset of the bank, and my liability. 2/5
When the transaction is complete, I have a new asset (house) worth $250k and an offsetting liability (mortgage) of $250k. The bank now has a new asset (mortgage loan) of $250k and an offsetting liability (deposits) of $250k and you now have $250k (cash) as an asset. 3/5