Could we use programmable capital to quantify a risk-free rate proxy?

You won’t get very far into the study of economics without somebody bringing up the theoretical idea of a “risk-free-rate” (RFR). This concept transcends any technology, legal or political construct. The crypto community is not immune to the theory and its implications.

“Currently there exists no benchmark discount rate for crypto assets. This makes it very challenging for investors to build out valuation models in the space.”

— Jon Todaro

Background of the Traditional Financial System

There are many issues with the current Risk-Free Rate. In practice, the best we can get is a society, industry or community-level consensus about which proxy to use. It’s boring stuff, but necessary. In the US, in some circles, the 3-month Treasury Bill is often used. In others, they choose the 10-year. Can that rate really be annualized? Is the US Government really free of risk? What happens as certain countries hover near negative rates? Is there a sliver of risk associated with USD? Academics have already tried beating this topic to death.

Crypto Credit, No Risk-Free Options

There has been a cambrian-explosion-like event in the last year or two, enabled by blockchain technology surrounding innovative new lending products. Dai, Dharma, Celsius, Compound, Nexo, ETHLend, just to name a few of the cool kids. All of them are experimenting. Some of them are starting to reach scale. The investors in those communities, whether they know it or not are pricing the risk involved with those loans. Parallel markets are forming. Depending on the architecture of the technology and market microstructure, if you sliced the risk premium many of those loans include risks such as:

1. Smart Contract Logic Risk — The risk that the programmable capital has an unintended error.
2. Smart Contract Exploit Risk — The risk that the programmable capital has a potential exploit which is not caught or known by audit.
3. Borrower Credit Risk — The probability the debt will be serviced.
4. Collateral Risk — The risk the collateral will fall in value.
5. Liquidation Risk — The risk that the collateral liquidation system will function as expected.
6. Stability Fee Risk — The risk that the Dai stability fee will change.
7. Currency Risk — The risk of non-local currency volatility.
8. Inflation Risk — The risk that inflation rates will change. Note this is different than an inflation premium, which is a market’s attempt to price current expectations.

When you account for these risks, in a probabilistic way, a risk premium emerges — which is added on top of the RFR, and any other premium (liquidity, inflation, etc.) to get an accurate yield.

The more interesting half of the crypto community has no elected government, and therefore no authority to issue risk-free bonds, to see what rate it trades.

The Ask

So, I ask the community — should we make a risk-free proxy? Should we run an experiment, similar to how the community effectively experimented with ENS, where we all kind of said: “Let’s try exploring this, and see what we learn”? Would the builders who follow, benefit from the insights we might gain? The code we might write? The market data which is revealed?

How could we do this? Programmable Capital.

If the community created a pot of money then sold that pot at auction, we would gain a proxy for the RFR. Imagine the headlines if the ETH RFR was proven lower than the US Governments.

The business logic of the smart contract would include 3 phases:

1. Create a block-time like rolling term structure which effectively gives us, say, 12 discrete pools of money. At T=0, these pools might be labeled month-13, month-14, month-15…month-25.
2. Allow the community to donate for a period of time to the pools of their choice. At the end of the donation period, say 12 months (T=12 months), the pools would become 1-month, 2-month…12-month.
3. Sell the rights to the pool of capital at auction. The winning bidder gets their own capital locked up for the same term, plus the capital which was donated to the pool.

This would create a series of instruments that would form a 12 month yield curve. That yield curve would reveal a risk-free-rate of crypto.

Optionally, it would be nice if a secondary market could form but I haven’t figured out how to guarantee it would.

Eliminating Risk

In practice, we could only get to a point of near-risk-free. Like all current proxies. All of the above would have to be done with code that is free of any governance, or even need for governance. Free of any middle-men. And audited extremely heavily. All actors involved would need to be completely altruistic in their pursuit for it to work. There is a subset of academics and a subset of open source devs who operate only for the sake of curiosity.

Who would chip in?

I would. I’d kick in to see what happened. To be part of the pioneers helping to bring transparency to models for the rest of the defi community? Hell yeah! Think of the financial liberalization which might be accelerated by the resulting data. I also imagine certain foundations, think-tanks, and sponsors might see value in this experiment. Anybody who has ever funded research about the risk-free-rate might participate. Basically, curious rich people and institutions. Maybe a badge, or something, is earned for any donation to the pools. Maybe we sell tongue-in-cheek T-shirts, “I helped quantify risk free crypto”.

How much is necessary?

If we found a way to raise even $1M, and the market bid on this, that could create something on the order of a $50M to $200M worth of notional — depending on the rate. Do you think there is $100M worth of crypto out there, which would bid on the pool? I do.

Who’s with me?

I have sufficient finance and technology knowledge to lead and execute this alone. However, given my other commitments in life, this would effectively be just a hobby. So if I do it alone, it might take a while — plus, it likely won’t work unless the community is supportive and ready to help.

All I want to know is…

Do you think this is a valuable idea?

Let me know. A clap on this article, or a star on this repo would signal sufficient feedback for me to push forward. Also, feedback on reddit is welcome too — please assume I’ll be lurking.

Note: I need to confirm with my employer that this isn’t a conflict with our policies somehow.