lukestein’s avatarlukestein’s Twitter Archive—№ 6,333

                  1. Suppose you had a side hustle catching snails on the weekend and delivering them to a restaurant catering to French tourists. You’d love to get some sense each Saturday of how many tourists are likely to arrive over the next week so you know how much time to spend snailing. 🐌 1/ @paulgp/1236122318730125313
                1. …in reply to @lukestein
                  You’re in luck: It turns out that there’s an arcane organization (the Coalition of Bakery Oven Examiners, or CBOE) who publish just such a forecast, called the FIX (i.e., the “French visitors IndeX”). Their forecast isn’t perfect, but it’s on their website so 🤷‍♂️. 2/
              1. …in reply to @lukestein
                What does this have to do with baking? Who knows and who cares?! You’ve got your forecast and it helps you with your snailing. But then you get an idea for a new side hustle. The French go on vacation in August. The FIX will go up then. You can predict it in advance! 💰 3/
            1. …in reply to @lukestein
              But… you can’t make money predicting the FIX, because: The FIX is not a *price*. It’s a forecast. It’s a number on a website. You can’t buy it or sell it or hold it. So you’re back to snailing. But then you get curious; how does the CBOE come up with their forecast? 4/
          1. …in reply to @lukestein
            It turns out the CBOE explains exactly how they do their forecast cboe.com/micro/vix/vixwhite.pdf . It turns out there’s a baker who makes the most amazing croissants 🥐 and baguettes 🥖 and the French tourists love them. 5/
        1. …in reply to @lukestein
          It turns out that every weekend this baker prints his menu for the upcoming week. If he expects lots of French tourists, he prices 🥐🥖 high; if he expects few, he prices low. He prices to maximize profit given his demand forecast. And he’s pretty good at forecasting! 6/
      1. …in reply to @lukestein
        So what does the CBOE do? Well, it turns out that their FIX index is just (the squareroot of a slightly adjusted) weighted average of the price of this bakery’s croissant Q(🥐) and baguette Q(🥖)! 7/
        oh my god twitter doesn’t include alt text from images in their API
    1. …in reply to @lukestein
      “Ah ha,” you say. “FIX *is* a price, it’s just the price of a particular bundle of 🥐+🥖. So now I know how to make money from knowing that the French vacation in August! All I have to do is buy lots of baked goods in July when they’re cheap and then hold them for a month!” 8/
  1. …in reply to @lukestein
    Of course this doesn’t work. Even though FIX *is* a price, it’s not a price of something durable. *This week’s* FIX is the price of *this week’s* 🥐+🥖. 9/ @lukestein/1233622273246015488
    1. …in reply to @lukestein
      Well, the VIX is (sort of) a forecast of market volatility. But it’s really (the squarerroot of a slightly adjusted) bundle of financial options contracts. Just like 🥐🥖 are expensive when French tourists are anticipated, options are expensive when volatility is expected. 10/
      1. …in reply to @lukestein
        For 🥐🥖, this is baker’s profit maximization. For options, this is no-arbitrage/stochastic calculus. And just as the baker might price high/low for reasons unrelated to 🇫🇷 tourist forecasts, option prices aren’t just determined the vol forecasts. @bennpeifert/1235684257881665537 11/
        1. …in reply to @lukestein
          And thus ends the weird #econbakingvolatilitytwitter thread that @paulgp gets for posting a weird gif at a time when I’ve been thinking about vol and how to explain it. 12/12
          1. …in reply to @lukestein
            A few bonus tweets about the real VIX published by the real CBOE. First off, VIX approximately forecasts the annualized standard deviation of returns on the S&P 500 index over the next 30 days. You can convert to monthly by dividing by √12 or daily… 13/ @lukestein/1053021398963298306
            1. …in reply to @lukestein
              The VIX does actually forecast S&P 500 volatility; i.e., the market tends to have bigger swings (up and down) when VIX is high. 14/ @lukestein/1236104906190151680
              1. …in reply to @lukestein
                There are vol indices for things other than the S&P 500 over the next 30 days (VIX). @CBOE has other horizons ($VIX9D $VIX3M $VIX6M) and other equities ($VXD $RVX $VXN $VXO). You can get vol indices on oil, or even the volatility of volatility $VVIX! 15/ @lukestein/1071160197958127616
                1. …in reply to @lukestein
                  Volatility tends to go up when the market is falling. (The timing is tricky, but you can think of this very roughly as “stocks go up slowly and down quickly.”) So investing in volatility can serve as a hedge (i.e, an imperfect insurance policy). BUT… 16/ @lukestein/1236090433974882304
                  1. …in reply to @lukestein
                    …just like trying to “invest in FIX” would involve a portfolio of 🥐🥖 that are constantly going stale, “investing in VIX” involves a portfolio of options that have 30-day maturity today, but only 29-day maturity tomorrow. 17/
                    1. …in reply to @lukestein
                      For this (and other) reasons, “investing in VIX” involves constantly selling slightly stale options and buying fresher ones. On avg, THIS IS A MONEY-LOSING strategy! But that’s OK; on average I lose money on my car insurance too. This is a portfolio insurance premium. 18/
                      oh my god twitter doesn’t include alt text from images in their API
                      1. …in reply to @lukestein
                        Much more in threads from @bennpeifert. He seems to write explainers on the train; start here @bennpeifert/1228479496405180416 , and if you want more, just search twitter for “train from:bennpeifert”. You may also recognize Benn’s name from his @QJEHarvard with @aprajitmahajan😉. 19/
                        oh my god twitter doesn’t include alt text from images in their API
                        1. …in reply to @lukestein
                          And if you want to play with daily data in @Stata, you can download from @stlouisfed (which has several vol series fred.stlouisfed.org/categories/32425 ) using eg freduse VIXCLS SP500, clear or from yahoo finance using e.g., getsymbols ^VIX ^GSPC, yahoo fy(2000) frequency(d) 20/
                          1. …in reply to @lukestein
                            So I edited this a bit and posted it (on LinkedIn, but you don’t need an account there to view or print cleanly to PDF) if that’s your jam @lukestein/1237167437490999296
                            1. …in reply to @lukestein