These HF traders are not investors they are skimming money from us real investors-who invest in companies. To stop this insanity thers should be a Transaction Tax on every trade. This tax would be miniscule, but would discourage HF traders.

Be ahead of our time. Once only computers do trading, it is a matter of time before they fight to a draw. Profit for either side is no longer possible at this point. Then what? Perhaps the "Haves" have all, or perhaps an insight occurs at this point. Why not be ahead of our time and have the insight now. As Bob Marley sang: "Doesn't justice stand for all mankind?" Let's get a jump and live the "tie" now. We all matter. Who should "win" the trade anyway?

One thing that was missing about the HF trading piece was the damage it did to the average person, like myself.
I had a few stocks that made up most of my savings and as (I thought)a prudent investor I had sell stops on all of them. The idea was that I could ignore the day-to-day movements of the stocks but if one of them got into trouble it would automatically sell thus minimizing my lose. I had he stops set at down 10%
What actually happened was that they all sold, some for much less than down 10% and I lost a great deal of my savings. Of course they then rebounded a few seconds later thereby transferring my savings into the hands of HF traders.
Lesson - If you're in the stock market you are a fool. Its rigged.

The race for information and trading speed advantage reminds me of the 19th century competition between the Philadelphia and New York stock exchanges. New York had the advantage of being nearer by ship to London. Some ships carried business information from London, and New York traders acted on that information first. The Philadelphia exchange attempted to blunt the advantage with a system of lighthouses, which conveyed information by signal, from New Jersey to Philadelphia. While the time advantage was diminished, it couldn't be eliminated, and eventually led to New York's domination of financial trading.

Wait until Quantum Computing gets into this mix!
BTW the cable length trick was used on the CRAY supercomputers computers (bundles of wires stuffed in the base of the towers.)

So we've got an arms race: if you can shave a millisecond off your propagation delay, you have an extra millisecond to conduct trade before the other guy gets the news. Which means a lot of what economists call "rent seeking" expenditures, a classic source of market inefficiency.

It seems to me that, if they wished, trading markets could largely end the arms race by introducing a random delay -- say, with a mean of 20 milliseconds -- into the processing of every trade. The random delay would have the same distribution for everybody, so it wouldn't affect the balance of power. But a 20-ms head-start doesn't make you much more money than a 1-ms head-start, so it's not the _mean_ that matters so much as how often you beat the other guy. And your odds of beating the other guy are basically 50/50, dominated by the random delay rather than your cable.

Your graphs! Label them!

Two things. First, no mention was made of 'stop orders'
a stop order is an order to sell a stock as soon as it drops to a certain price. The idea is that you minimize the amount you can lose, but the problem is that you are only guaranteed to put in a sell order, not guaranteed to actually trade at that price. Anytime you put in a sell order you automatically change the price to the highest corresponding buy order. If there are way more sell orders than buy orders, you can inadvertently drop the price to an irrationally low amount. This can set off even more stop orders. The result is an acceleration of stop orders being executed and an enormous drop in price as fewer and fewer buyers step in to stabilize the price. There were multiple stocks whose price dropped to essentially zero in milliseconds simply due to stop orders.

Second, there was a little bit of blurring of the differences in types of trading. High Frequency Trading is a subset of algorithmic trading. HFTs typically do not hold a position for longer than 24 hours. Essentially, their algorithms seek out arbitrage opportunities and are not speculative in nature. Other algorithmic trading IS speculative in nature and positions can be held for indefinite amounts of time. Both types are based on the traders' interpretation of market information and their success is ultimately based on the skill of whomever creates the algorithm. An algorithm's ability to react to price changes in microseconds does not imply that an HFT is behind the algorithm. There is another type of trading called 'statistical arbitrage' which in theory locks in a profit by taking advantage of the difference in price between KNOWN orders. This can be possible when there is an inefficient clearing process. Statistical arbitrage could be classified as a type of HFT. Some people feel like traders that engage in statistical arbitrage are essentially gaming the system at the expense of other market participants.

WHY is this allowed to happen?

WHY is there no transaction tariff assigned to trades, especially the frequent ones?

In terms of science, can fractal geometry be applied to the stock market?

Please don't show graphs without at least some attempt to explain them! What is the video actually showing? What do the colors represent? What are the x and y axes on the graphs? There is no science here, just "dumbing down" ("it is pretty and it looks like fire"???)

Excellent show, gents!

The last part (using extremely cold sodium to slow or stop light) made more questions in my mind than answers. Doesn't Newton's law apply to light (can't remeber if it's law 1, 2, or 3, but the one about an object in motion staying in motion and an object at rest staying at rest unless acted on by an opposite force)? I get that the cold sodium acts as an opposite force to stop or slow the light, but the scientist interviewed says that if it manages to get through the Sodium, it speeds back up. What causes it to accelerate again and why doesn't it continue moving at it's slow speed, once it's slowed down?

This is both fascinating and deeply disturbing. As I watch these graphics, I experience a sensation as if I'm catching a glimpse inside the "matrix" or seeing the lifeblood flowing through the throbbing "heart" of some huge, monstrous beast.

Nice show. It reminded me of an experience I had when I was graduate student in Neurobiology at Purdue University:

During one hot summer day some friends and I went to a flooded quarry where we jumped off a cliff of about 40 feet into some deep water. I noticed that I was going so fast when I got to at that level. I mentioned this to my friends (mostly science grad students) and they were having a similar experience.

This I attributed mostly to synaptic delay at the synapses between the photoreceptor and the visual cortex. At least two synapses in the retina (such as: photoreceptor to bipolar cell, bipolar cell to ganglion cell) plus others on the way to the cortex.

This is something where anyone willing to jump off a tall cliff into water can easily experience the slowness of their nervous system functioning themselves.

Blowing my mind was an understatement.

Dave from England...I think you articulated your concern very well. I figured more people might have found this very disturbing. I haven't done a lot more research but I thought I'd share first thing I ran across:

http://www.cnbc.com/id/49333454

About high-frequency trading.

I am not sure I can articulate this properly, but doesn't the "arms race" for speed of trades point to the fact that there is something very wrong with the stock market?

I don't see how you can "make" money without there being a corresponding change in the real value of something. If there is no real change in value, then for some people to make money, other people must lose money. The worrying thing is the high percentage of trades that are under the control of automated systems, with complexities that no-one understands. This would all be fine if the fall-out was only limited to people directly involved in trading on the market, but global economies are affected, and thus ultimately everyone.

About the pitch drop experiment.

In my current job, I debug computer hardware problems, we often have to debug a failure that does not occur at a high frequency. None of them are quite on the scale of happening only once every 8 years, but I think the techniques we use, can be of some help.

In order to make progress, we like the failure to happen at least once a day. So if the failure naturally happens. Once a week, then we setup seven systems doing the exact same thing. Which means we should have one failure a day. If you setup 8 of these pitch drops, you should have a drop every year.

Another way of speeding up the failure is by modifying the environment to force it to occur more often. Many times changing the temperature will change the frequency of the occurrence. If you put the experiment into a chamber and heated up the experiment, you would see the drop happen more frequently.

Great segment, this one on the speed of the trade market. I missed more details about the algorithms used to trade - if a few feet of cable can make a big difference in the speed of trade, surely a few more lines of code or more efficient computations could also have an impact. Did your interviewee say anything about that?