Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?
Back for a limited time - Get 15% off sitewide on Slashdot Deals with coupon code "BLACKFRIDAY" (some exclusions apply)". ×

Comment Re:where is the random? (Score 1) 395

I think the situation people mostly worry about with market making is where Alice submits an offer to sell for at least $1.00 and Bob submits an offer to buy for no more than $1.02. There's $0.02 of surplus here, and we would expect it to get somehow split between Alice and Bob. When we add in Martin (market maker) to the scene, people worry that he would accept both Alice's and Bob's offers and pocket some of that $0.02 surplus for himself. If Alice and Bob both put in their orders at the same time, then Martin is obviously left out, but it's never really simultaneous. This leaves a time window when Martin could potentially insert his own pair of buy and sell orders. If it's a rather slow market, and Bob's orders appears a few days after Alice's, it's reasonably clear what service Martin is providing (Alice can quickly and reliably offload this thing she no longer wants, and the item is readily available for Bob to buy when he finally wants it).

On the other hand, if the market isn't so slow, and Martin is doing HFT, it's less clear what Alice and Bob gain from him. If he's able to take a position and close it in a matter of seconds, people start to wonder just how much better off Alice is from having been able to finish her sale seconds earlier. It looks like we've got diminishing returns as Martin's holding periods shorten. His claim that he's providing liquidity looks dubious in a market that has to already be pretty liquid to support the activity people see from him, so they suspect they're being hoodwinked. Modern technology enables really large scale market making, so Martin's got a lot of money flowing in, which makes people even more suspicious.

Personally, as a rather small-time investor, I don't really fret about this. Even if it is the case that I am getting some of what would otherwise be my money skimmed off by someone claiming to have provided liquidity that was already present, I don't make enough trades for that potential loss to matter much. I worry more about whether this is a good (for society) use of the math/engineering talent we produce.

Comment Re:Hansen again? (Score 4, Informative) 605

and then research a bit about sun activity, sunspot-cycle

Now show that this warming trend is really just the upward half of a fluctuation that's been repeating every eleven years.

Oh, you didn't know the sunspot cycle was only eleven years long? Maybe you should have researched a bit about sun activity.

Comment Re:Dangerous move (Score 1) 182

Seems to me that for an internship to be educational to the intern, she needs to be doing something useful.

IIRC, a SCOTUS case dealt with a trainee program run by a railroad company where prospective rail workers would essentially do simulated work, supervised by actual rail workers. SCOTUS ruled that their arrangement did not require pay, and listed the fact that it was simulated, not real, work as a necessary condition.

Comment Re:If your working, then yea, that should be paid. (Score 1) 182

The question is whether they really want to be unpaid or if unpaid work is just something they have to bend over and take in order to have a career. If the system effectively requires those new to the field to work without pay for a time, it is certainly not for the new workers' benefit, yet those defending the unpaid work keep on claiming that it is.

Comment Re:NASA isn't good at listening (Score 1) 319

The fact is the current Soyuz launch record without fatalities is quite definitely and literally significantly better.

Except the "current streak" metric is close to meaningless. It depends too much on when the sample is taken. Even with two agencies of equal success rates, it is very unlikely for them to have equally long success streaks at an arbitrary point in time. This is the same reason why "current uptime" is not a good metric for system stability. Run the numbers for MTTF, and then we have something to talk about.

Comment Re:proofreading for the college graduate? (Score 1) 836

Is learning another language really so hard for you? With a good background in CS, it shouldn't take you more than a week to be able to start producing code in a new language. If you can't keep up with the arrival of new tools, you probably missed something while you were in school.
The skills you claim are valuable would not help you here; the skills you claim are useless are practically required.

Comment Re:Algorithms (Score 1) 836

Good programmers should be able to pass a competency test for any employer. If they pass that test, they should be able to seek the job, degree or not.

The process HR has in place (including checking the applicant's education) saves a lot of time (and thus money) in exchange for a relatively small hit to accuracy.

Quark! Quark! Beware the quantum duck!