This (or the 95/5 version you mention later) makes a lot of sense with things like Bitcoin.

Say you're on track to have a nice retirement. You put 5% or 10% of your portfolio into Bitcoin. Worst case, it goes to zero and your retirement will be at worst 5% or 10% less affluent. I say "at worst" because if it goes to zero, any remaining investments will not go into Bitcoin so your more conservative portfolio will grow a little faster.

Best case is hard to assess, because Bitcoin could go up an annualized average of 10%, 50%, or 100% over the coming years. If we assume (dangerous, but inescapably you have to make some sort of guess here) Bitcoin doubles on average every four years, as a result of the halving rate, and you keep your money there for the next 12 years, it'll go up 8x.

Comparing that to the plausible ~2.25x increase of stocks/bonds over 12 years (based on a 7% annual increase), that 5% or 10% of your portfolio will lead to an extra 25% increase in your overall ending portfolio value.

Here's the calculation:

$1M invested at 7% annually over 12 years -> $2.25M

$900k invested at 7% annually plus $100k invested at 19% annually, all over 12 years -> $2.825M

$2.825M / $2.25M = 1.255 -> 25.5% greater final result.

Assuming the same with a 5% speculative position gets you $2.539M, or 12.8% larger ending portfolio balance.

Of course, this assumes Bitcoin increases on average just under 19% a year. If crypto bulls are right and it increases 100x over the coming 12 years, a 10% allocation to BTC will net you an extra $777k or 34.5% overall portfolio increase.

A 5% allocation using the same return assumptions will net you an extra $390k or 17.3% larger ending balance.

In summary, this is a case of assymetric risk/benefit. If you can stomach the risk of a posiition going to zero that will cost you 5% or 10% of your expected nest egg to open the possibility of a 12.8% or larger increase, this sort of speculative allocation makes sense.