As somebody who used to work in Program Trading, I suspect that unanticipated side effects of Reg NMS amplified already substantial market movements. The regulation, designed to prevent trade-through, has the effect of creating a submarket for those interested in fast execution at any cost. This submarket is very light on liquidity.
It seems conceivable that everything functioned exactly as it was designed.
As the Yen spiked, holders of much borrowed USD needed to liquidate and pay back their borrowings, and program selling snowballed across very very thin markets.
Several 'liquidity providers' including TradeBot withdrew liquidity in the plunge, perhaps leading to the 0.00 and 0.01 bids that applied to quite a few stocks.
We have a very poorly regulated system, with few safeguards for regular investors. Sure, the spreads are generally tighter than the old days of floor trading, but the secret dark markets and computer trading can be turned on and off at will, leading to the kind of thing we saw Thursday.
You seem to be under the erroneous assumption that human specialists and market makers actually caught falling knives in the past. Sure the rulebooks said they were supposed to, but rules aren't reality.
Black Monday is the classic example. Lots of phones were ringing and simply not getting picked up because they didn't want to deal with the orders. That's even worse than today. Unless you were on the right trading floors, you couldn't know what was happening and you couldn't cancel or enter new orders.
Someone called it a "display of digital nukes". Here is a quick chart I did on spot market EURJPY which is a proxy barometer for equity markets (it is corelated, like copper price is for world economic growth): http://i.imgur.com/E6MK2.png massive inverse cup and handle on a daily chart. So it is by TA, straight from a textbook. Movement was rather dramatic though, and still more to come after a bit of consolidation (from a week to about end of the summer at most).
But you have to keep in mind, the market is specifically afraid of the Euro collapsing, so it might not indicate anything today other than the fact that people would rather have JPY than EUR.
EUR/JPY is a known gauge of market sentiment, same as copper is for economic growth. Markets go up, EURJPY goes up, markets go down, EURJPY goes down. It's a known correlation that hasn't been broken yet. Intramarket analysis is a widely popular field in the economics. Just because people are afraid it doesn't mean it won't happen. Every possible TA is indicating on further collapse to come this year, at most by the end of summer. Even Elliot Wave suggests it if you are into EW. Even fundamentals are all wrong, carry unwind just happened, dollar is surging because of the credit market collapsing, vix is going mad, ecb is incompetent/inert. ECB will try to do something, tomorrow I'm sure - but it's a bit too little too late. Avalanche had its first run this week. We'll see a period of consolidation for a week, to up to a summers end if we are lucky! People can downvote my first post, but they cannot ignore that chart and should not.
"Past performance is not indicative of future returns."
A large country in the Eurozone has not defaulted on its debt before. This default affects the price of the Euro. You can't just look at a graph and assume that it's always right.
The Euro isn't even old enough to make claims like "always", anyway. It's only been on the market since 1999 and only in circulation since 2002.
My naive guess is automatic stops. If you enter a position electronically, you set stop losses for the low barrier. If it falls below that, the stops will trigger, selling off the stock, lowering the price, and causing other stocks to do the same creating a domino effect.
But there are people on the other side of the equation, too, with orders to buy when the price reaches a certain number. And there are the options traders.
