Monday

This past week, starting Monday, Jake Bailey and I conducted a one-word survey:

How much Bitcoin will you buy with your stimulus check this week?

Our goal was to get 2,000 surveys completed (“completes”) before Wednesday when the stimulus payments hit. This would give us a statistically-significant sample of US citizens.

Wednesday

By Wednesday morning, we had 200 completes. Not enough, but went ahead and posted the results anyway in a few tweets:

Jake Bailey (no twitter) and I are running a survey to see how much Bitcoin people will buy with their stimulus checks.

We have 340/2000 completes done. Here are the results so far: (thread)

— Hunter (@huntermmonk) March 17, 2021

Of the people who qualify, 6.41% will buy BTC. pic.twitter.com/BXzqDoe0di

— Hunter (@huntermmonk) March 17, 2021

Google estimates there are ~328M US citizens: https://t.co/U9KrUpXb7q

In 2019, about 53% of Americans were making 75k or less (the stimulus cutoff): https://t.co/CCUww1cXoM.

— Hunter (@huntermmonk) March 17, 2021

Leaving us with an estimated 175,915,200 people receiving the stimulus. (rough numbers)

If we assume that in each of the buckets, the person buys the middle amount (if they answer $101-$250, on average they buy $175.5), we get the amounts: pic.twitter.com/gLLzJe8Cut

— Hunter (@huntermmonk) March 17, 2021

That comes to a total of $6,332,947,200 of BTC people plan to buy with their stimulus money.

For comparison, according to https://t.co/vlVdmMjeAy, daily volume for the past 24 hours was $6,549,258,827 (about the same). pic.twitter.com/Ve1taNNrry

— Hunter (@huntermmonk) March 17, 2021

Trade Result

Even though the survey wasn’t completed, it seemed pretty positive, so we went ahead with the trade anyway.

Wednesday morning we made the purchase. Added some buying and selling through the day.

Price immediately moved.

Arrow indicates when the tweets were sent.

We ended with a 0.79 BTC profit ($45,302).

Post-mortem

It’s unclear if the price change was related to stimulus or not. Although it was profitable, there were a few problems with our execution.

1. We started with only ~200 completes. I believe that means we had a 6.9% error rate.
2. Our data was self-reported.
3. We used Google Surveys, which will skew towards higher income individuals with internet access. That may or may not be salient given all Bitcoin purchasers are internet users.

And then, after execution, we saw that 24 hour volume was pretty normal which did not fit with our thesis. We expected that the influx of money would significantly increase volume.

I missed taking a screenshot here, but 24 hour volume was in the 6-8 billion range according to bitcointradevolume.

Also, as of writing, we now have 1,805 of the 2,000 completes.

People have received their check during the survey which skews the results. Nevertheless, here are the current results:

Total intent to purchase ends up being pretty much the same; definitely within the error from the beginning of the survey. ($500M more than before)