The article basically suggests that algos are reducing market volatility.
Algos don't try really hard to get VWAP. VWAP is REALLY easy to achieve, _beating_ VWAP is what most banks are trying to do when they trade large blocks.
A lot of the stories about warring algos and detecting each other are just that: stories. Yes, it has happened, but by and large thats not the game thats going on.
In the extreme case, consider you were on the other side of every other trade that happened in the market. You would by definition have achieved VWAP over some time period.
Trading randomly during any given time period probably comes close to achieving VWAP. I'm sure something a little more intelligent could probably achieve it in expectation.
VWAP is "volume-weighted average price," not "average price". Therefore, to replicate it, you need to participate in line with volume-at-time, not 1/N per time bucket. This requires you to estimate the volume envelope over the time period of interest. (Often a day, or in the case of hedging certain new issue convertible bonds, up to several days.)
Yeah, that's why I said in the extreme case you have to participate in all the volume. Of course randomly trading doesn't prove that you achieve VWAP, but I'm just saying that it's probably not terribly far off.
Many brokers will guarantee an execution at VWAP. It's a competitive business. The last time I put in such an order, I traded a double-digit percentage of the day's volume and paid 10bp commission (0.10%).
Brokers guarantee VWAP only if they also are paid a very healthy commission. They generally lose to VWAP, but makeup for it in premiums.
Plus, if you are dealing with a large broker they may just cross your order with an offsetting order (or their own position, execute nothing in the market and keep the entire premium.
It would be interesting to know how much of this 25k/day trade is simply a kind of steady-state speculation, including attempts to exploit arbitrage. If somebody really wanted to change 100k BTC to $, it would likely have a disastrous impact on exchange rates.
The trade of US$ in the currency markets is on the order of several trillions per day. Now imagine what would happen if somebody truly wanted to exchange a few trillion US$ for Euros, and keep those Euros...
The global currency post market turnover is ~1.4 trillion according to the BIS, note this includes many currency pairs that dont involve USD; a trillon a day is an upper bound on the daily USD market.
China effectively traded around 0.2 trillion USD for yuan and kept the USD last quarter, the impact is not that big.
More generally looking at the volume transacted does not tell you much about the depth of the market; you need to look at the order book for that. Looking at the bitcoin market there are several lots of over 25k sitting within 20 cents on each side of the bid-ask, this means it is extremely likely one could carry out a directional 100k trade with relatively small market impact if spread out over a few days.
Thank you for the data on the Bitcoin market, that is indeed informative.
As for China though, I would like to point out that those trades must be seen in relation to the massive net flow of goods from China to the US (and other parts of the world). The trade you claimed is actually bigger than Chinese net exports for that quarter (which I find somewhat surprising, I have to admit, but probably that just balances out with what was happening in the past), but it's in the same order of magnitude.
If China did not do such trades regularly, the Yuan would be expected to appreciate in a noticeable way, making current trade arrangements more difficult to maintain. Essentially, if China did not make such trades, it would be forced to restructure its economy towards domestic consumption, and it would force the rest of the world to start producing more of their own stuff again, or swallow the price increases.
So to claim that the impact of those trades is not big is problematic.
Can someone explain the velocity of money argument to me?
The blog post seems to be:
1. If btc are liquid, everyone will want to move btc wealth out to some other store of wealth and btc will fail.
2. If btc are illiquid, then it fails by definition. Currently btc are at this stage.
Here's why I don't get it:
1. If they ARE liquid and its easy to convert btc value to other value, why would everyone want to move value out of btc? seems like people would want to move value IN to btc because of the other benefits (anonymity, instantaneous transaction). Regardless, it doesn't follow that easy value conversion predicates value drain.
2. "Bitcoin seems to be at this stage [of being unable to convert Bitcoin assets into other, non-Bitcoin assets easily now]" -- this seems to be false. There's a highly liquid market for BTC-USD (21k btc have traded so far today). Yes, its not easy to transact 500k, but thats true for most new assets including exchange listed backed equities.
sorry, I wasn't clear. I wasn't asking WHAT velocity of money is, I was asking someone to explain the author's velocity of money argument. i.e. transfering wealth out of btc accelerates the velocity.
q, m, v, and p remain relatively constant in that transaction unless the liquidity event causes a persistent decrease in the price level (no evidence that the current BTC market could support it, but also c.f. above my comments about immature markets and liquidity).
My reading of his argument is that bitcoins primary function will be as a medium of exchange, not as a store of value. Thus when any indiviudal has a significant amount of bitcoins, he will convert them to an asset class that does store value.
nothing. anyone can. right now its mostly inefficient to allocate cpus for mining. its just not worth it for them. as more bitcoins are mined they get harder to mine.
something like that. This morning I had this thought for 2 things:
1. a python filtering function I wrote
2. how to restructure an entire django project.
Both (esp #2) are a little big for posting a code snippet. Bigger architecture problems probably only make sense within the context of the larger scope, so I'd pretty much want someone to check out the whole repository.
The SO thing would be great because SO has built in reputations so it'd help people feel more comfortable with giving away the entire project. I personally don't care, but I can imagine that being a barrier to code reviews by strangers.
Obviously, that sort of code review is a much more time-expensive request, so finding reviewers might also be hard without some sort of compensation.
I'm not super concerned about confidentiality for my code, but then again I'm not doing anything groundbreaking.
Mostly, if I cared about confidentiality, it'd be so I don't get publicly embarassed for crappy code.
Rating systems for reviewers would also let people build reputations so that reviewee's could feel more comfortable about code confidentiality.
Stack overflow already has reputations built in, so as I mentioned, it might just be awesome as a feature for SO. And another revenue stream, if they weren't already swimming in cash.
The main confidentiality issue is not about whether you, the coder, cares about people stealing your groundbreaking idea. When you are doing work for hire, the code is not necessarily yours to publish.
Whilst sharing it with someone you could refer to as a colleague, employee, or contractor might be OK. Sharing it with the world, or someone who might share it with the world, is not.
There's a balance between deliberately writing garbage and trying to start with perfection, and you choose some middle ground.
Yes, write maintainable code. No, don't accrue technical debt just for the fun of it. The post is titled as if all other things equal I would prefer paths that result in more technical debt, but that's clearly moronic.
But you have to start with step 1 (or I guess you can start with step 2...). The point for me is that I end up a lot more long term productive if I get my ass in gear on "get it done" first, because it helps me get to "test" and "refactor" a lot sooner.
That doesn't have to be true for everyone, but it seems to be true for me... this week anyways.
From the first footnote on that blog post, this has got to be the best non-computer hack I've read so far:
One of the Y Combinator questions asked you to name one non-computer system that you’d hacked in some interesting way. My answer concerned a man-in-the-middle attack I once did on Craigslist personals. I placed an ad as a woman seeking a man, and as a man seeking a woman, and then simply crossed the email streams by forwarding mail from one to the other, and vice versa. Most Craigslist personals didn’t even have photos back then, so the switch went undetected, even after the couples had met. I handed off the relationship by telling one that the other’s email address had changed, from my fake one to the real one, and likewise vice versa. For all I know, those couples are still together and having kids. They probably don’t know to this day what happened or what brought them together.
I hate to be a killjoy but neither party ever asked why the e-mail address was changed resulting in the other one jumping in and saying "I never changed my email address, you changed yours!!"? And thus opening up an investigation as they show each other the emails from the past. Secondly, neither party ever asked "Was this the first time you posted an ad on craigslist?" and finally neither party ever said "I don't usually respond to ads on craigslist but..." (people say this out of insecurity not because they actually mean it, this is just to make themselves look good).
I know this because I use to do this all the time in college over skype , record long awkward conversations between distant aquaintences and roll on the floor laughing daiwa try to figure out who called whom and who they are. This was actually a lot of fun, it was an exercise in psychology, we would try to predict how crazy explosive weird conversations could get and see what happens based on people's personality profiles.
My point is that they would have to be really really stupid to not know that somebody else connected them...
The article basically suggests that algos are reducing market volatility.
Algos don't try really hard to get VWAP. VWAP is REALLY easy to achieve, _beating_ VWAP is what most banks are trying to do when they trade large blocks.
A lot of the stories about warring algos and detecting each other are just that: stories. Yes, it has happened, but by and large thats not the game thats going on.