110% agreed. This is a near-textbook example of "just because you can doesn't mean you should".
The simple reason in this age of high frequency trading and massive amounts of money being moved around, a tiny arbitrage opportunity could instantly be magnified and trigger a system (market) crash. Whereas these crashes can and do occur today, the non-realtime aspect of settlement and clearing mitigates to a large extent. See the 2010 flash crash example: https://en.wikipedia.org/wiki/2010_flash_crash
In addition, I'm dubious of how "real time" this proposal will turn out to be. For one there are a number of dependencies and steps that need to be fully aligned to expedite from end of day to something faster. True real time systems are a rarity - most of the time the latency is still measured in minutes or hours. A flashy term that tech companies like to throw around.
Also this feels like a diversion tactic from the implosion of Robinhood of the company from last week's debacle. For those unaware, RH also put limits on AMD stock to be traded - what the heck?!
> Whereas these crashes can and do occur today, the non-realtime aspect of settlement and clearing mitigates to a large extent. See the 2010 flash crash example: https://en.wikipedia.org/wiki/2010_flash_crash
How so? I'd argue that the SEC policy of breaking "clearly erroneous" trades before settlement exacerbates that kind of flash crash: market-makers can't step in to prop up prices and then hedge their exposure, because their trades will be broken and their hedges won't. If settlement was real-time, preventing flash crashes would be profitable and people would do it.
> Also this feels like a diversion tactic from the implosion of Robinhood of the company from last week's debacle. For those unaware, RH also put limits on AMD stock to be traded - what the heck?!
It's not a diversion, it's a genuine explanation - the collateral requirements were why RH had to impose a bunch of limits.
> I'd argue that the SEC policy of breaking "clearly erroneous" trades before settlement exacerbates that kind of flash crash: market-makers can't step in to prop up prices and then hedge their exposure, because their trades will be broken and their hedges won't.
I'm a bit lost on your response. Are you advocating against SEC or stock exchange intervention? Remember that trading can be halted by stock exchanges also in order to decelerate massive movements. If the settlement happens on even a by-minute increment, then there is no effective recourse for these interventions. Flash crashes are black swan events and as such there's little incentive (as of now) for companies to build safeguards against it.
> It's not a diversion, it's a genuine explanation - the collateral requirements were why RH had to impose a bunch of limits.
It's an explanation but doesn't explain why other stock brokers didn't impose the same limitations, nor why RH is tapping its credit lines. I say it's a diversion because it's an attempt to deflect blame for its own poor business practices and policies - e.g. if my server can't handle the load, I can't just say "well, if I had a faster load balancer...." to my users.
> "It's an explanation but doesn't explain why other stock brokers didn't impose the same limitations"
Other brokers did impose trading restrictions on GME (and other stocks) around the same time as Robinhood. Webull, M1 Finance, Public, and E-Trade all halted buys of GME.
Others, including Interactive Brokers, TD Ameritrade, Schwab, and Trading 212 implemented restrictions such as halting options trading and greatly increasing margin requirements.
> Are you advocating against SEC or stock exchange intervention? Remember that trading can be halted by stock exchanges also in order to decelerate massive movements. If the settlement happens on even a by-minute increment, then there is no effective recourse for these interventions.
Yes, I'm advocating against them. My position is that SEC and exchange intervention makes flash crashes more rather than less severe. (Also they're harmless; artificially suppressing the volatility of the stock market makes everything more fragile, we'd be better off embracing them and making sure all market participants are equipped to handle volatility. It's like how decades of suppressing forest fires has made forest fires much worse)
> It's an explanation but doesn't explain why other stock brokers didn't impose the same limitations, nor why RH is tapping its credit lines. I say it's a diversion because it's an attempt to deflect blame for its own poor business practices and policies - e.g. if my server can't handle the load, I can't just say "well, if I had a faster load balancer...." to my users.
If your hosting service suddenly cuts your server capacity down by 2/3, I think it'd be fair to complain about that to your users.
According to Robinhood, they didn’t want to put limits on anything, it was the cleaning house that made them, so there had to be something that triggered AMD restrictions?
It looks like that because it's complex and technical and people don't understand it, but it also saved retail investors a bunch of money. I don't understand how a jet engine works, but I don't feel any urge to advocate for jet engine policy changes.
For those who don't understand your comment about HFT saving money for retail investors, several well-regarded published studies suggest that the narrowing of mean bid-offer spreads since the introduction of HFT can be attributed to HFT.
There are actually a variety of HFT strategies, but there's good evidence that the high-speed market-making variety are good for small volume investors. Those in favor of getting rid of HFT should be specific about exactly which sorts of HFT they want to ban.
The narrowing of mean bid-offer spreads isn't surprising, since many HFT strategies are essentially faster versions of traditional market making. They're able to offer narrower spreads (better prices) than traditional market makers because they're faster at widening spreads (worsening prices) when the market starts to move in one direction. On average, this means better prices for people trading small amounts of stock. However, for large institutions that trade market-moving volumes, low-latency market making means worse prices, since the market makers are faster at noticing and widening spreads when there are large market-moving trades afoot.
Also, the ETF arbitrage variety of HFT strategies reduce the amount that ETF prices differ from the prices of the individual stocks within them. This results in fairer prices for people who primarily invest in ETFs.
On the other hand, low-latency market-making results in less liquidity being offered when the markets are moving quickly in one direction, which is when it's most important to have liquidity.
Also, for the average person, a very large portion of their exposure to the stock markets is through pension funds and mutual funds, which tend to make the sorts of large-volume market-moving trades that are hurt by HFT's ability to notice and quickly get out of the way to avoid being run over.
Disclosure: I've never worked for an HFT fund, but I do work in the financial services sector. I did some work on systems designed to reduce market impact of large institutional trades, part of which is breaking up patterns so that HFTs are worse at noticing and widening their spreads.
The simple reason in this age of high frequency trading and massive amounts of money being moved around, a tiny arbitrage opportunity could instantly be magnified and trigger a system (market) crash. Whereas these crashes can and do occur today, the non-realtime aspect of settlement and clearing mitigates to a large extent. See the 2010 flash crash example: https://en.wikipedia.org/wiki/2010_flash_crash
In addition, I'm dubious of how "real time" this proposal will turn out to be. For one there are a number of dependencies and steps that need to be fully aligned to expedite from end of day to something faster. True real time systems are a rarity - most of the time the latency is still measured in minutes or hours. A flashy term that tech companies like to throw around.
Also this feels like a diversion tactic from the implosion of Robinhood of the company from last week's debacle. For those unaware, RH also put limits on AMD stock to be traded - what the heck?!