The 0.20% is if you win no prizes at all. On average including prizes, the EV on an annualized basis is around 3% right now. We are subsidizing some of the prizes currently, which is why it's higher than other HYSAs. In the long run we will aim to match Marcus and Ally, which right now are at around ~1% but go up or down depending on market cycles.

I appreciate the response, I've got a few remaining questions that perhaps you can shed some light on:

- Is that only for the average or also for the median?

- Is the data published anywhere (would love to see distribution curve)?

- Are the calculations assuming someone is saving every week? Or has it also been tested monthly / bi-monthly?

- When someone gets a "windfall", do they keep it in their account or do they withdraw it? (i.e. no longer savings but spent in some sort of celebration)

- Is this using an actual random number generator in which case there will be long-tails of people winning much more as well as losing much more? Or is there some sort of engine that ensures everyone gets something?

- This is the average. Going along with your second question you can get a sense of the distribution by looking at the payouts and probabilities in the official rules (https://www.withyotta.com/official-rules)

- It's purely probabilistic so it doesn't need to be tested. It's just random number math.

- Wins are deposited into your Yotta account and are immediately available to withdraw if you'd like

- It's an actual random number generator

It looks like the current base rate is a very competitive yield, and I'd expect the odds to be lowered soon, since their lottery gives the equivalent of about 3.6% in interest in addition to the 0.20 base interest. Another startup (Level Bank) using the same banking partner (Evolve Bank and Trust), launched with 2.1% APY in February, but soon had to lower it to 0.5%. This comment (https://news.ycombinator.com/user?id=adammoelis) by the CEO suggests they are subsidizing the lottery temporarily.

The variance should be relatively low for deposits in the 5 figures and above. See the results of my calculations below, based in part on the spreadsheet linked from https://news.ycombinator.com/item?id=23781247

If one holds $10K in the account, one should get 400 entries a week, and the expected number of times to hit each prize under $1 is between 1 and 9 times every week. I calculate the equivalent interest rate from the <$1 prizes alone is about 2.02%.

The $10 and $15 prizes should be hit about once every 2-3 months, and add about another 1.16% to the expected winnings.

Since the $300 and up prizes are split amongst all winners, it's hard to calculate the expected value, but assuming 10 million entries a week, the EV of the larger prizes is adds about 0.42% extra per year, with the vast majority coming from the expected value of the $5.8 million prize.

I believe that would be very unlikely for one to get below the equivalent of 2% interest over a few months if they held tens of thousands of dollars in an account, with the median outcome being around 3.4%. One could run a simulation to confirm this.

> Unless you have math that says the average winnings people will get will make up for the .7-8% difference you see other institutions giving?

That is how I personally interpreted this statement from the OP:

> We provide a rate of return on savings that on average is in-line with the top yielding savings accounts out there like Marcus or Ally.

If that is true, then you're either choosing to earn your ~1-2% interest rate slowly but surely (in one of the high-yield savings accounts you listed) or in "batches" (using Yotta). It makes sense to me why people would prefer the latter. It also makes sense to me why many people here on HN would prefer the former.

I don’t see it as a disadvantage at all.

If people who are already saving in a high yield savings account move all of their funds to this, that’s almost entirely on them.

On the other hand, if this entices people who aren’t saving at all to start saving, that’s a huge advantage for them.