I'm pretty excited about the bitcoin prediction markets. It's a nice way to coordinate distributed knowledge about the future -- and have people put actual, real-world money behind their projections.
This one, though, seems outrageous: Will the U.S. National Debt Cross $40trn Before July 4th?. It probably just opened and got seeded, which is why the odds are so stupid (even).
I pretty much see no way this is going to happen.
Some general observations:
  • We're at $36trn today.
  • Fiscal Year 2024 (year ended Sept 2024) deficits = $1.83 trillion
  • U.S. nominal GDP about $29.1trn
  • Last month the Treasury ran a $257bn deficit (Fiscal Data); these things come in unevenly but even if you extrapolate that over the next eight months (=2.056trn) you're still nowhere near the 40trn mark.
  • if we take the highest single-month deficits in recent years ($380bn) and extrapolate that, we get about 3trillion -- again, roughly a trillion short.
  • Fed is bringing BS down, not up, meaning the appetite for opportunistically issuing more Treasurys that the market could stomach is limited.
Looking at this in a different roundabout way:
  • U.S. growth might be 2.5% next year (Goldman Sachs)
  • U.S. deficit should be around 7.6% of GDP (Fitch)
  • That gives us 29.1 x 1.025 x 0.076 =$2.27 trillion worth of deficits (for _the entire FY 2025, not just until July 4th).
If NGDP growth were 5%, pretty extraordinary as we're hinting at recession etc, and the deficit blew out to 10-15%, only then would we get close to the mark for the full Fiscal Year 2025.
Even with some Trump tax cuts early next year (which probs won't pass Congress etc., in time to do much) there's no way they're running 2-years plus worth of deficits in eight months. And the Elon-Vivek DOGE thing, if likely to do anything at all, is on cutting expenditures, thus operating against the debt blow out to 40trn.
While I don't particularly like locking up sats for this long a period (the election gamble where most of us made out like bandits was a good trial run, but also over a much short time horizon), this seems like a no-brainer.
Free sats on the table. Will I grab some?
194 sats \ 5 replies \ @Arceris 5h
Regarding the election odds, I’m a bit confused about the payout.
I dropped 5000 sats on 10/13, on the Who Wins the NY race. I recall the odds were atupid that the Democrats would win (like only 50-60%). Looking at the timestamp, I’m one of the earliest bets, so it should have been in that insane range.
But I only realized 5630 (after the vig) which seems a bit low for winning a wager in that range. Shouldn”t it have been closer to 8k or 9k? It would be helpful if the betting record recorded the implied probability of the bets.
reply
It doesn't. The implied probability of the bet is for that moment, means nothing for later payout (as I'm starting to learn).
It's just the sum of the pots divided across winner and loser (minus fee)
reply
17 sats \ 3 replies \ @Arceris 3h
Well, good thing I didn't go big. I'd be rather annoyed that they're using parimutuel odds and it isn't like a prediction market and I'd have gone big based on the implied probability.
reply
Yeah we're working on an LMSR version where you have fixed odds when you place a bet, and you'll be able to exit your position if the event outcome looks unlikely (hopefully out by EOY or early next year). We went with the parimutuel system because it was simple to develop, but it's by far not the best for incentivizing early bettors or for long-term event forecasting.
reply
exactly. Good to just try these things out with smaller amounts
I would have made lots more money from the Trump election bet had I known about this and wasn't concerned with the counterparty/payout problems that someone (@south_korea_ln ?) listed above
29 sats \ 1 reply \ @Shugard 6h
Honestly, without thinking at all, I thought "yeah crystal clear we will have more than 40 trillion, government is stupid!"
Reading this, I'm like, "Fuck! I'm already too much of a pessimist.
So are we going to bet on no and hope for all the other uneducated bitcoiners like me?
reply
Totally. I'm taking suckers' corn ALL DAY LONG, BRO!
reply
Unless they changed their system (they said they would), one should not bet so long in advance on https://www.bitcoinprediction.market/. You can wait until the last minute, you'll avoid any risk (especially with this kind of bet) with the same profit as if you had bet today.
reply
yea, that's the portion of this I haven't quite understood yet. If I don't get the odds at betting time but merely split whatever pot the losers put in (minus fees) there is, as you say, no point doing things too early.
The 50:50 I'm seeing right now, isn't the payout I'm getting
reply
reply
nice, thanks!
Yea, what I don't like about that obviously is that I'm not just betting the outcome of the bet but also one level up, i.e., how will others on the platform bet this bet. Keynesian beauty contest on top of an actual real-world outcome.
Ugh.
reply
We're working on changing the system. LMSR version should be out by end of year or early next year!
reply
36trn only? How much trns have been added in the last four years? Trump “contributed” with like 8 trns or similar figure during his administration - yeah, this Trump who had nothing to do with the inflation whatsoever. Is it possible to bet on the us debt in four years.
reply
nope, this is a myth.
It was even my original MONEY CLASS! #733580
reply