3 ms·

During the 2008 crisis Buffett bought a $5 billion stake in Goldman Sachs. He didn't buy shares on any exchange, he dealt directly with Goldman. Buffett doesn'

During the 2008 crisis Buffett bought a $5 billion stake in Goldman Sachs. He didn't buy shares on any exchange, he dealt directly with Goldman. Buffett doesn't need complex financial markets because he does custom transactions directly with companies. A Winklevoss could easily sell him a billion-dollar five-year put, but he wouldn't like the price.

isolli 9y ago
I think he would worry more about counterparty risk.
JumpCrisscross 9y ago
> he wouldn't like the price For fun, let's try pricing a Bitcoin put option using the Black-Scholes-Merton model [1]. This model is a simplified version of real-life options models, but our parameter uncertainty will dominate the model's anyway. I'll use the $14,400 Coinbase price [2] for both our S and K inputs, i.e. we're pricing a European put struck at spot. Term is 5 years. This site [3] says Bitcoin's historic 252-day volatility is 5.21%. Sure. We want the 5-year 365-day volatility, i.e. 14.0% [4]. The real shit show in this example is the rates component. If we use this 57% number [5], the put is worthless. We're using a Bitcoin-dollar price for the stock and strike, so let's assume U.S. dollar funding (since Bitcoin does not natively support rates in the way modern currencies do). The 5-year Treasury yields 2.25% [6]. BSM says...$1,037.01. (Rho dominates tau, or in English, the rates component is about 2.5x more meaningful than the volatility component.) Raising our rate to the call money rate [7] we get $792.84 with rates continuing to dominate volatility. [1] https://en.wikipedia.org/wiki/Black–Scholes_model https://en.wikipedia.org/wiki/Black–Scholes_model [2] https://www.coinbase.com/charts https://www.coinbase.com/charts [3] https://bitvol.info https://bitvol.info [4] sqrt([5 x 365] / 252) x 5.21 [5] https://www.bloomberg.com/news/articles/2017-11-09/keynes-worked-out-the-bitcoin-interest-rate-it-s-57 https://www.bloomberg.com/news/articles/2017-11-09/keynes-wo... [6] https://www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=yield https://www.treasury.gov/resource-center/data-chart-center/i... [7] https://www.bankrate.com/rates/interest-rates/call-money.aspx https://www.bankrate.com/rates/interest-rates/call-money.asp...
- bluedevil2k 9y ago
  You didn't finish pricing out the put option. I calculate $475 for a put. But think about this a second. That's a crazy low price for this kind of investment. You're only paying a 3.3% premium to hold a put I can exercise at any point in the next 5 years. At any point in the next 5 years, if BTC goes down by rougly 7% I'm in the money. That's a daily swing for BTC in a normal week. I think the volatility input in this equation is way off.
  - bluedevil2k 9y ago
    And anyone who thinks this is a fair price for a PUT on BTC, please contact me as I will buy unlimited amounts of puts if you're willing to put up 50% as collateral.
    - FabHK 9y ago
      Me too! Depending on your credit, that is. Finally a get-rich-quick-scheme I can get behind.
  - JumpCrisscross 9y ago
    > I think the volatility input in this equation is way off I agree. One's ability to price a put on Bitcoin is probably anti-correlated with one's willingness to sell one.
- FabHK 9y ago
  I'll buy a 5yr ATM put at 1038 USD! 1. There's no way rates dominate vol for BTC. 2. Bitcoin vol is way higher than that. It's obviously higher than most stocks, and stocks are in the 20% range. My quick calculation gives me an annualised vol of about 90% over the last year, and 140% over the last month. That gives a put value between USD 8000 and 12000, in other words approaching the "infinite vol" limit (present value of strike). EDIT to add: BSM takes annualised vol (or, to put it differently, vol enters only in the term sigma^2*T, and that product must be dimensionless, thus the unit of vol is 1/sqrt(year), or whatever unit you use for T). To compute it, take the average of squared log returns over a period (ie, take average of LN(P(t)/P(t-1))^2), annualise by multiplying with 365 if you have daily prices (or 252 if you have business daily prices), and take the square root of that.
MarkMc 9y ago
2008 was an exception - most of Buffett's investments have been in the open market where he was simply the highest bidder at the time. The problem with a Winklevoss put is likely the credit risk, not the price

isolli 9y ago

I think he would worry more about counterparty risk.

JumpCrisscross 9y ago

> he wouldn't like the price For fun, let's try pricing a Bitcoin put option using the Black-Scholes-Merton model [1]. This model is a simplified version of real-life options models, but our parameter uncertainty will dominate the model's anyway. I'll use the $14,400 Coinbase price [2] for both our S and K inputs, i.e. we're pricing a European put struck at spot. Term is 5 years. This site [3] says Bitcoin's historic 252-day volatility is 5.21%. Sure. We want the 5-year 365-day volatility, i.e. 14.0% [4]. The real shit show in this example is the rates component. If we use this 57% number [5], the put is worthless. We're using a Bitcoin-dollar price for the stock and strike, so let's assume U.S. dollar funding (since Bitcoin does not natively support rates in the way modern currencies do). The 5-year Treasury yields 2.25% [6]. BSM says...$1,037.01. (Rho dominates tau, or in English, the rates component is about 2.5x more meaningful than the volatility component.) Raising our rate to the call money rate [7] we get $792.84 with rates continuing to dominate volatility. [1] https://en.wikipedia.org/wiki/Black–Scholes_model https://en.wikipedia.org/wiki/Black–Scholes_model [2] https://www.coinbase.com/charts https://www.coinbase.com/charts [3] https://bitvol.info https://bitvol.info [4] sqrt([5 x 365] / 252) x 5.21 [5] https://www.bloomberg.com/news/articles/2017-11-09/keynes-worked-out-the-bitcoin-interest-rate-it-s-57 https://www.bloomberg.com/news/articles/2017-11-09/keynes-wo... [6] https://www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=yield https://www.treasury.gov/resource-center/data-chart-center/i... [7] https://www.bankrate.com/rates/interest-rates/call-money.aspx https://www.bankrate.com/rates/interest-rates/call-money.asp...

bluedevil2k 9y ago
You didn't finish pricing out the put option. I calculate $475 for a put. But think about this a second. That's a crazy low price for this kind of investment. You're only paying a 3.3% premium to hold a put I can exercise at any point in the next 5 years. At any point in the next 5 years, if BTC goes down by rougly 7% I'm in the money. That's a daily swing for BTC in a normal week. I think the volatility input in this equation is way off.
- bluedevil2k 9y ago
  And anyone who thinks this is a fair price for a PUT on BTC, please contact me as I will buy unlimited amounts of puts if you're willing to put up 50% as collateral.
  - FabHK 9y ago
    Me too! Depending on your credit, that is. Finally a get-rich-quick-scheme I can get behind.
- JumpCrisscross 9y ago
  > I think the volatility input in this equation is way off I agree. One's ability to price a put on Bitcoin is probably anti-correlated with one's willingness to sell one.
FabHK 9y ago
I'll buy a 5yr ATM put at 1038 USD! 1. There's no way rates dominate vol for BTC. 2. Bitcoin vol is way higher than that. It's obviously higher than most stocks, and stocks are in the 20% range. My quick calculation gives me an annualised vol of about 90% over the last year, and 140% over the last month. That gives a put value between USD 8000 and 12000, in other words approaching the "infinite vol" limit (present value of strike). EDIT to add: BSM takes annualised vol (or, to put it differently, vol enters only in the term sigma^2*T, and that product must be dimensionless, thus the unit of vol is 1/sqrt(year), or whatever unit you use for T). To compute it, take the average of squared log returns over a period (ie, take average of LN(P(t)/P(t-1))^2), annualise by multiplying with 365 if you have daily prices (or 252 if you have business daily prices), and take the square root of that.

MarkMc 9y ago

2008 was an exception - most of Buffett's investments have been in the open market where he was simply the highest bidder at the time. The problem with a Winklevoss put is likely the credit risk, not the price