The hard part about analysing whether is actually is good is in trying to figure out just what value we should place on added liquidity.
Can you explain why the value to society of the additional liquidity from HFT outweighs the cost (large profits to companies doing nothing else)? Not being facetious, actually interested in hearing your thoughts on it.
Don't buy into the HFT "adding liquidity" myth.
I still can't figured out a single value-add they bring to the table, much less to justify their insane compensation or the incredible technology budgets they carry. It's simply sad to me to see some of these brilliant minds (at least on the tech side where I interact) essentially engaged in ripping off fellow citizens at worst and rent-seeking at best.
Of course if you engage them, they will get all hand-wavy about "liquidity" - but I think it's striking that not a single person I've talked to has been able to distill the benefit they bring to society in a simple sentence or paragraph. It always ends with the "most people can't understand" which as I get older I've learned is a euphemism for either "I don't know" or "I don't want to tell you the truth".
Either way, so many billions of dollars taken from the economy and put into useless activity. Absolutely no one will be able to convince me we need millisecond-level liquidity for markets to function efficiently.
The complaint about "brilliant minds being wasted" rings true, but you could make the same complaint about all of finance, and even most of Silicon Valley focused on extracting rents by serving ads - both are ultimately zero-sum games where insiders profit at the loss of society.
I think the ultimate problem is, there are no real problems that smart people could work on and make a contribution (and earn a good living) - or at least I don't see any.
I don't know the details of HFTs in general, but to my understanding, what they do is arbitrage (orthogonal to liquidity) - basically, making sure that any simple patterns in the price are exploited (and thus reduced). Note, however, that HFTs themselves don't make money - the markets could be just as efficient if only HFTs would participate, and with any new piece of information (earnings release, news, movements of other prices) a new equilibrium would establish (best bid and best ask) without any trades happening. Trading either comes from (1) disagreement about the price - and the person who's right earns money, rightfully IMO (and if you don't want to lose money, just don't play), or (2) external circumstances (e.g. you want to buy a house => you need money => you sell your stocks), in which case HFTs are actually helping you (reducing the spread, and still earning money - e.g. a "stupid market maker" offers you 100 and then sells it for 101, earning 1, whereas the "smart HFT" offers you 100.01, and then sells it for 101, earning 0.99 - and you're better off as well).
This isn't the whole story, of course - there's probably a lot of illegal things going on (dark pools, front-running, order flow selling), and a lot of waste could be reduced (e.g. by regulators forcing HFTs to share/rent infrastructure and only compete on algorithms, or mandating everyone has the same delay to the exchange), but those are problems only if you let HFTs exist.
Now things are a bit more complicated because HFT is a fairly broad term but this is true in a broad sense.
Without commenting on this particular barrier to trade, it's certainly the case that in the past, some nations have put up barriers to trade that have worked out very well for them.
I like you explaination, but exchanges are nowadays mostly private companies so the 'tax' goes to shareholders and not to the state.
This is just another fluff piece without too many rational arguments, based on a completely irrational fear of HFT. Arguably, HFT and hedge fund have made the markets much more micro-efficient in the past decades (as wittnessed by the difficulty of finding new arbitrage opportunities); we have to thank the central banks and big institutional investors for macro-inefficiency.
> one that applies a tiny tax rate to an array of transactions and is split between buyers and sellers
I think the current situation, where exchanges basically "tax" those who take liquidity and reward those who provide it (i.e. the fee of a market order is split between the exchange and the person who entered the limit order) is much better at providing stability to the market.
> which generate windfall profits on small and fleeting differences in prices at the expense of ordinary investors and market stability
Well, the idea behind is that prices (e.g. same stock on different markets) should converge. If you don't want that, by all means, tax the market.
> The burden of this tax would be concentrated at the top, because that’s where the ownership of financial assets is concentrated
Nope, it doesn't matter where ownership is concentrated. What matters is where trading is concentrated. E.g. HFT or every time you get paid and invest a portion in your income in a stock index-backed pension scheme.