You can find a varying number of practitioners and academics on both sides of
this debate. To be honest, the question of whether "High Frequency Traders"
increase liquidity is ill-posed. The label is often misused and is broadly
encompasing of too many different types of traders.
So, in general:
Any trader that posts resting limit orders is adding liquidity to the
Any limit order resting on the order book is immediately executable by any
trader with sufficient liquidity demands.
Passive traders attempt to limit when they trade by canceling their resting
orders when they believe a change in price is probably. Not doing so
exposes their orders to adverse selection.
When a passive trader's orders are filled, especially when adversely
selected, they sometimes, depending on their specific circumstances, have
to immediately demand liquidity from the market place to reduce or limit
The above applies to all participants, whether they are "high frequency" or
You can find firms that are primarily aggressive takers of liquidity. You can
also find firms that are primarily passive traders posting limit orders. This
is true of "HFT" firms all the way to individual traders.
Many detractors believe that passive HFTs do not add high quality liquidity to the
market place. They point to high cancel/fill rates and other characteristics
of their order flow as evidence. In my mind this represents a type of "model
bigotry". These traders are simply better than many other participants and are
using more sophisticated models. Without official market maker obligations a
HFT, indeed ANY trader, should be free to manage their orders in whichever way
they wish. If it means their liquidity is too hard for you to access, then you
as a counter party must either choose not to trade with them or become better.
I'll also add, some people even go so far as to suggest that aggressive
traders also add liquidity to the market place. After all, limit orders won't
be matched unless someone is willing to cross the spread and demand liquidity.
This is a point of view that likely differs significantly from the academic
viewpoint of liquidity. I first heard it from Jaffray Woodriff of QIM fame,
and have come to believe it to be true.
So, to answer your question directly, what is the basis for the argument that
a "high frequency trader" increases liquidity? I'll provide two answers:
Any trader, whether a HFT or not, that posts orders is adding liquidity.
Any trader, whether a HFT or not, that takes liquidity is "adding
liquidity" to the market place by trading with a resting limit order.