flash orders is actually the main thing i was referring to when i said "There are flavors of HFT that aren't easily defensible "
a whole lot of HFT has nothing to do with seeing the orders before they happen.
"high frequency trading" encompasses a lot of different strategies (just like "computers" is a large and varied field).
imagine a high frequency strategy that finds mispriced securities, and trades them. as it happens, these trades tend to converge very quickly. would you consider this "fundamentally different" from long term investing (in any way other than the holding time horizon, which the agent can't possibly know before the trade actually realizes)?
Yes, I would. The definition of "mispriced", I would argue, is quite different. You could come up with all sorts of algorithms to trade in the midst of a bubble, and find "undervalued" companies on micro time scales. In the value investor's mind, the same companies are always overvalued. Value is relative in one case and absolute in the other.
Splitting hairs perhaps, but the fundamental approaches are different. I will give you that the hf mentality can certainly be applied over long time periods.
Where you and i diverge is "Value is relative in one case and absolute in the other."
I think you're saying that HF Guy's definition of value is relative and Value Guy's definition of value is absolute.
If its even correct that any company has a single true value, I think this number is unknowable. Investors or traders make estimates of this value, and if their estimate deviates significantly from the market price, they transact with Mr. Market.
Nobody's estimate is more absolute or relative than any one else's.
Comments
HFT is not the same as long term investing. Being able to see orders before they happen makes it fundamentally different.
flash orders is actually the main thing i was referring to when i said "There are flavors of HFT that aren't easily defensible "
a whole lot of HFT has nothing to do with seeing the orders before they happen.
"high frequency trading" encompasses a lot of different strategies (just like "computers" is a large and varied field).
imagine a high frequency strategy that finds mispriced securities, and trades them. as it happens, these trades tend to converge very quickly. would you consider this "fundamentally different" from long term investing (in any way other than the holding time horizon, which the agent can't possibly know before the trade actually realizes)?
Yes, I would. The definition of "mispriced", I would argue, is quite different. You could come up with all sorts of algorithms to trade in the midst of a bubble, and find "undervalued" companies on micro time scales. In the value investor's mind, the same companies are always overvalued. Value is relative in one case and absolute in the other.
Splitting hairs perhaps, but the fundamental approaches are different. I will give you that the hf mentality can certainly be applied over long time periods.
define: mispriced = "value != market price"
Where you and i diverge is "Value is relative in one case and absolute in the other."
I think you're saying that HF Guy's definition of value is relative and Value Guy's definition of value is absolute.
If its even correct that any company has a single true value, I think this number is unknowable. Investors or traders make estimates of this value, and if their estimate deviates significantly from the market price, they transact with Mr. Market.
Nobody's estimate is more absolute or relative than any one else's.
edit: deleted a bunch of pointless words.