3 ms·
>Capitalism cannot work without risk premiums/cost of capital/hurdle rates. Still works: As someone who was throwing pennies in front of the credit risk steam
by cinquemb 7y ago
>Capitalism cannot work without risk premiums/cost of capital/hurdle rates.
Still works: As someone who was throwing pennies in front of the credit risk steam roller since 2018 (long OTM puts on junk bonds with 4-12months of expiry, roll every 2 months), people were gladly picking up,
I would give my theoretical first born for markets to go back to ATH so I can make a bigger pos on long on tail risk… helps when writers/dealers were (and still are) pricing options with black scholes horrible assumptions for OTM and when not near expiry…now they gotta buy back at +20x with help from the discount window (shout out to head of the 3 time felonious JPM: Jamie Dimon, hope for a swift recovery… id have heart problems too if my TBTF bank was at the discount window).
Mispriced risk is amazing, and cheap to be long while playing other high risk (high prob payoff, short tail risk) strats until it pays off.
These fake interest rates set by the FRBNY bureaucRATS with their daily manipulations in coohoots with the +20 brokest-dealers, only make tail risk cheaper to buy.
- H8crilA 7y agoI agree and I also have some Taleb-like tail hedges. My point, though, is that no risk premiums doesn't work for the economy, not for you or me. It causes bad companies to exist. It causes bad projects to be funded.
- cinquemb 7y agoBut that's the thing, there are risk premiums, but most of the time their not priced in way that reflects the risk "accurately" because the writers don't think they'll have to pay out (and as along as most insurance works by calculating payout probabilities and collecting payment as income, there will be systemic incentives to massage probabilities of the unlikeliest events, rather than say, treating it as short term loan where you can derive income from lending part of it in money markets while still covering payouts, and returning the value/rolling the coverage for the next interval). And sure, it might enable companies to survive longer if premiums were higher, but "bad" companies/projects can still come into existence because its not obvious ahead of time that if choosing at random (or even if you made an better than random guess, you cant eliminate the possibility), which particular company will be a "bad" company, or a "bad" project.