,
11 tweets,
3 min read
Read on Twitter
Theories will first follow #crypto prices (attempting to explain), but then prices will follow the theories that achieve mental mass. placeholder.vc/blog/2018/11/1…
A theory on valuation will gain recognition by having "data-model" fit on *past data* & then being decently good at predicting future data.
If the theory is good enough, the majority of asset allocators will come to use it, *bending future data* into it (prices follow theories)
If the theory is good enough, the majority of asset allocators will come to use it, *bending future data* into it (prices follow theories)
I can't emphasize enough how much of a mind-bender the last bit is, because it also means the best theory may not win.
Instead, it will be the theory that gets to enough mental mass first such that it starts to bend future data into its model, and thereby becomes predictive.
Instead, it will be the theory that gets to enough mental mass first such that it starts to bend future data into its model, and thereby becomes predictive.
Prices are, after all, an entirely human creation, crafted out of the intelligence (or lack thereof) of our collective models and assumptions.
Lastly: I think only "fundamental valuation" models will be powerful enough to bend future reality/data in the way I've explained, which is why they're the holy grail for #crypto analysts.
"Relative valuation" models will remain useful but are not specific/convincing enough to warp the future imo. They tend to build upon the warped reality put in place by a "fundamental valuation" model, comparing an asset to itself across time, or comparing an asset to its peers.
Oops, not done - a model that gets to mental mass may be displaced if a gross error is revealed, or if underlying values shift. For the latter, look at how DCFs for equities have gone through many contortions, depending on if earnings, or EBITDA, or sales, or ______ are valued.
It's not to say DCFs have been displaced, but the way a DCF is used to value an equity has become massively heterogenous and changed siginificantly with time.
Should add a caveat here. A better model may come about later and still “win” if it is clearly superior (per the math) to the point that it is able to overcome the “bending of the numbers advantage” that the prior theory has.
Good further thought, somewhat to this end, here:
