David Beckworth is pushing for more creation of safe assets.
“One of the big challenges facing the global economy is the shortage of safe assets. These are the highly liquid, information-insensitive assets that function as money. The financial crisis raised the demand for these safe assets just as many of them were disappearing. This cyclically-driven safe asset shortage (or excessmoney demand) that emerged continues to this day and is why aggregate demand in many countries remain depressed.1″ [italics mine]
Specifically, he would prefer the creation of safe private assets. I wrote about the links between money and collateral a few times in the posts on money and money-like instruments. And before we get going, remember it’s pretty clear we live under a real estate monetary standard.
However, there are two related problems with creating safe private assets which should be recognized.
- There is a limit on the creation of the information insensitive assets
- We can’t easily tell how close we are to this limit.
Most private safe assets are created using pools of real estate as the backing collateral. Here is a breakdown of the numbers from April. You can see only 11% of the April tri-party repo market uses something other than real estate or government debt as the collateral. Government securities underpin about 36% of the tri-party repo market. Real estate accounts for about 51% of the total. These percentages are for tri-party repo – there is a whole market for bilateral repo which appears to have a roughly similar breakdown.
The repo market uses either government debt or mortgage backed securities as collateral for a large majority of the transactions in the repo market.
We want a functioning repo market. We need to provide the repo market with safe private assets. Safe private assets are “informationally insensitive”. Informationally insensitive means “stable prices for these assets”.
And stable prices for real estate are exactly what the unholy bargin of monetary policy and real estate cannot provide in the near future.
The issue comes down to the weak arbitrage relationship between different methods of valuation, the zero bound, and time lag.
There are two different, semi-objective ways to value any real asset. You can use discounted cash flows, or you can use replacement value.
Discounted cash flows (DFC) simply takes the amount of cash this asset generates over the expected time frame and puts a value on that pile of cash. Since you get some of this cash in the future, you need to discount that cash to get an equivalent value of cash in hand today. The way to discount infinite cash flows going off into the future is to divide by the going rate of interest. So if you were going to get $100 per year forever and the interest rate is 5%, you just divide $100 by .05 and you get $2,000.
You can see the problem pretty quickly for the zero bound. If interest rates are 1%, that same $100 is now worth $10,000. if interest rates are .2%, that same $100 is worth $50,000. If interest rates are zero, the value is undefined. And if below zero – like they are today – then you end up with nonsensical valuations of this string of cash flows.
It seems to me as though real assets get their maximum value when interest rates are close to zero. This should cause fear in people, because we are close to zero interest rates right now and have been for many years already. It’s pretty clear we are close to the maximum value for real estate across much of the United States.
Still, there is another way to value real assets. You can value real assets through their replacement cost, or how much it would cost to replace or duplicate these assets. If trying to value a house, how much would it cost to build a near-duplicate from the ground up?
The two valuation methods don’t need to match perfectly, or even really match at all. It’s entirely possible for assets to generate low cash flows but be very expensive to replace. So by the Discounted cash flow method, the value would be very low, but by the replacement method, the value would be high.
On the other hand, it’s entirely possible for assets to generate great cash flows and be very cheap to replace or duplicate. So by discounted cash flows, the value would be very high, but by replacement value, the value is low.
We must remember there is a weak arbitrage relationship between these valuation methods, but only in one direction. When cash flows are high and replacement costs are low, there is a huge incentive for people to build new real estate assets. It makes sense for builders to go buy the raw materials and build new homes, because they are able to buy low and sell high, which means they make money.
So you can see how monetary policy works with real estate – it causes a disconnect between how much the cash flows from real estate are worth and the replacement value of this real estate.
This is what happened during the real estate bubble. Real estate generated nice cash flows, and building new houses was very inexpensive relative to how much cash they would generate.
It’s hugely important to remember building houses takes time, whereas revaluation by cash flows happens immediately or if you believe in expectations, even sooner! Changing the value of trillions of dollars of real estate can happen in a few minutes as interest rates change. Building trillions of housing stock takes years.
So there is a weak arbitrage relationship – something where people can make free money, but it just takes a long time for this arbitrage to happen.
As long as we were far away from the zero bound, we can continue to raise housing prices through lowering interest rates. Pushing down the interest rate allows valuation by discounted cash flows to increase immediately but does not cause the replacement cost to go up much, or as fast.
Once we get to close to the zero bound, we have problems. First, small changes in the interest rate cause huge valuation changes for the discounted cash flow method. This makes it unclear what a reasonable value is using DCF.
We know we are close to the maximum value, but not how close, because valuation seems arbitrary between using .2% interest rates and .5% interest rates, especially when these rates can change in the future.
Then, since we are close to maximum values, home builders want to exploit their arbitrage relationship between cash flow valuation (which is close to a maximum) and replacement cost valuation. They want to do this immediately, but this takes time – it takes lots of time to build houses. We know real estate transactions take a lot of time even after the house is built. So if you were a real estate builder, you’d want to under build houses until it was very certain you could lock in a huge profit – and then build as much as possible. But your lag time between planning/building and then selling can be very long. This is not a good way to create stable prices for real estate.
How can this result in stable prices for real estate? We have multiple issues for real estate:
- Zero bound makes DCF valuation difficult
- Weak arbitrage between DCF and replacement resultes in boom/bust mentality for builders
- Long build/delivery time to consumer means uncertainty for price setting
Now, at this point, remember the repo market wants information insensitive assets to use as collateral. They want assets which only change in value a small amount when interest rates change or NGDP changes. Yet, being near the zero bound causes a real world situation for real estate which makes information extremely important to the value of the assets used.
This isn’t a minor issue, and any amount of Quantitative Easing only makes it worse, not better.
Also, this analysis doesn’t even touch on the reduced importance of real assets to the overall economy as technology becomes more important. Izabella Kaminska has been great on this idea.