Basics of Quantitative Trading


Shaun Overton: So I got the impression
that thereís just a massive difference between where everybodyís at.
So youíre already actively involved in a Forex Market, youíve traded Options, youíve never traded before at all…
what about you? Attendee: Iíve lost so much money, my wife almost divorced me.
Shaun Overton: You see, thatís what I was talking about before you came in.
This conversation weíre having, itís very expensive. Uh, and then… Fan? 0:00:30
Attendee: Uh, I trade some stock and I trade forex, options stuff… Shaun Overton: Okay.
Attendee: Kind of like, getting in. I can program to buy-sell, so I want to see how you’re doing the programming
as part of the best strategy for trading. Shaun Overton: Okay, yeah, thatís kind of why
I didnít prepare a presentation as I wasnít sure where everybody was coming from,
so I didnít know what would be appropriate. So I think, really, what is best
is to start off with the high level problem for how you go about algorithmic trading,
and weíll do programming type of approach, ëcoz I think everybody in here codes to some degree.
So, you really just have a couple of very simple problems. You have… you need to get data,
you need to analyze it, and then you need to trade it. So, really simple on service,
but when you are trying to go about this, all of this can be huge obstacles.
So for example, letís say that I want to trade Forex data or I want to trade Forex
and I donít know anything about the Market, I donít know what softwareís common.
If I was doing this and I wasnít using regular software, I would have to go looking high and low on the internet.
And the quality of the data, you donít necessarily know anything about,
and you donít know where to even start looking. So if you are trying to do a free option,
you might go look somewhere like Yahoo! Finance where you can get data on lots of stuff that is mostly
end of day, but certainly high frequency that you can kind of get a sense for most securities,
you can get some options data, you can get Forex, um, you can get some indeces and some futures data,
not really all that much. Ah, you can get it also from Paid options,
but theyíre not really designed for Retail Traders and everybody in here sounds like theyíre a Retail Trader.
So, itís really… in the market thereís two types of people that weíre talking about,
thereís institutional; they work for hedge funds, they work for mutual funds,
they work for guys that manage $50,000,000 to several billion dollars.
And then thereís guys like ourselves that have regular jobs and then decide that they want to trade
and theyíre interested in algorithmic trading. So the Paid option is not really a good plan.
Uh, if youíre trying to buy data, usually it starts at several thousand dollars per instrument
and it goes up from there. Itís designed for Institutions
and itís just generally out of the Retail Traders league. Uh, and then finally, you can get it from the broker yourself.
Although, again, the quality of the data is very different. And the reason I keep talking about data quality
is that when we go on to analyze and come up with potential ways
that you might want to trade algorithmically, if you have garbage data, you have a garbage analysis.
So you have to be very careful about the data that you decide to accept.
Attendee: Is it faulty or delayed or…? Shaun Overton: Sometimes itís just wrong.
Um, sometimes thereís duplicate entries. Um, sometimes thereís gaps in the data that are unexplained.
And if you donít really know what youíre looking for, what kind of problems there might be in the data,
you can come up with some weird discoveries. Um… Attendee: The machine didnít parse it and all?
Shaun Overton: Yeah. Attendee: IBFX. Thanks.
Shaun Overton: Yeah, thereís, I mean… basically if you have, if you get free data, youíre getting what you paid for
and even if it comes from a reputable broker… Attendee: Yeah.
Shaun Overton: … itís still gonna have some problems. So you have to be very careful
before you go into all the cool quant stuff that youíve actually been careful with what you accept.
And then other times there are Platforms/APIs have all sorts of different companies that offer all this stuff,
all in the same product. So the advantage to those types of software is that it makes your life
a lot easier because if you go about this on your own itís a monumental task.
Okay. And then the analysis, Iíve heard all sorts of stuff, that thereís a lot of Python programmers in here.
Um, a lot of people that like to play with R, um, so you can use, I mean,
you can just custom program your own research platform. Iíve done that plenty of times
where I just create a little text file, then parse it, and then I start doing all the analysis in.
You can do R, you can do MATLAB, or then again, you can use somebody elseís platform in API.
And now, when it comes time to trade, if youíve done custom R or MATLAB or all that stuff,
you have the same problem where you have to build your own platform
or build within a platform, and then you have to have some kind of separate application
to start pushing through the trades. Or if you have a platform where everything is already written,
you can do this all in the same spot and it makes your life a lot easier.
So, the title of the chat is ìForec Columnî or ìPopular Forex Platformsî
and this is the option that weíre talking about. So it allows you to obtain the data without…
and weíll go through the counting odds, but you can generally believe
that the data is reasonable quality, if not solid. You can do all the analysis on the data
and you can place trades, and you can do all in one fell swoop without having to recode everything three times.
And so, in Forex, the most popular platforms are MetaTrader, you have NinjaTrader
and you have TradeStation. But this here,
I mean, I probably shouldnít have called this the Seminar, or whatever weíre calling it, Forex specific,
because most of these are popular with different groups. And then thereís some Brokers specific ones
that are really popular, ThinkOrSwim comes up most often. Okay, so these are just different software options
and thereís nothing special about them. Itís just what everybody happens to like
and is really common throughout the market. So, if you like trading Forex,
MetaTrader is overwhelmingly the most popular. You can literally find
more than a thousand brokers around the world that offer MetaTrader.
I mean, itís gigantic, you really have your pick. You can pick a different country,
a different size, liquidity, all that stuff. NinjaTrader is getting more popular with Forex,
but for the most part, itís FuturesTraders. So guys that like trading, ah,
the S&P minis, that like trading beans, coffee,
I mean literally anything futures. I donít know why but everybody loves NinjaTrader.
And then finally, TradeStation. They have their fair share of Futures Traders,
but for the most part, people who has the TradeStation trade equities.
So they, like in being often in the stock market and it makes it really easy to hook everything together.
And then ThinkOrSwim is really equities as well. So, I just want to check in with everybody.
Does everyone really kind of understand the platforms and why itís advantageous to go with a platform?
Attendee: Um… Multicharts? Shaun Overton: Yeah, Multicharts is
sort of the illegal cousin of the TradeStation. Multicharts is a Russian company and…
Attendee: I didnít know that. Shaun Overton: Yeah. Okay,
so all these platforms use their own different language. So MetaTrader uses a custom language called MQL4,
which is really a C Scripting Language. Okay, NinjaTrader uses C Sharp Dot Net 3.5.
TradeStation and Multicharts is a language called Easy Language, which for old programmers in here
will probably bore you to tears. Ah, itís super simple.
These two are not object oriented, and obviously, this one is. And Easy Language got popular
because it was not designed for programmers, it was designed for Joe Retail trader that is intimidated by programming.
So if you wanted to buy… letís pretend you have two bars.
This bar closed up about this bar, so you might say, if the close of one bar ago is greater
than the close of two bars ago, then buy at market. And then it has that English like syntax,
so itís totally not intimidating for somebody that doesnít have any programming background at all.
And this is all a long way of tying us back into Multicharts. Okay, so TradeStation got popular
because this language was so simple and it has a lot of cool features
that solve these three problems, the data analysis and hooking up to live trading.
So, what Multicharts did was, they basically just piggybacked entirely off TradeStation
and stole their entire platform, including their custom language,
and they offered it at a hefty discount. And now theyíve gotten more established.
They offer a lot of new features, but itís the TradeStation rip-off
that is slowly evolving into its own platform. Attendee: Okay, how compatible to… is the, is the,
ah, parallel language versus easy language? Is it a hundred percent…?
Shaun Overton: No. Attendee: …or 99?
Shaun Overton: Yeah, itís like 97, 98… I mean, youíre gonna break it.
So, I have developers, when we get an easy language project,
a lot of them will choose to work with Multicharts and then inevitably the client gets it and theyíre like…
ìWhat? You know, you didnít test this? What happened?î Then the problem is that these, these are very very similar,
but yeah, they are, there are differences. And especially as TradeStation adds new features,
especially like the GUI work and, um, anything that is more complex than this,
thatís where it just runs off the rails. Attendee: Great.
Shaun Overton: Okay, so Iíve covered the languages, the markets they cover and then, um,
the other problem that I mentioned is really data, so these platforms offer very different set-ups
and they all handle the problem differently. So, MetaTrader is kind of like the AK47 of Trade Platforms.
So, you download it and it doesnít matter how novice or not good with computers you are,
youíre gonna have a hard time not figuring out how to work the platform.
Itís stupid simple, itís very friendly, ah, itís also not sophisticated.
So if youíre trying to do something sophisticated, thatís probably not the place to be.
And for your analysis, you have to be very very careful, because when you download MetaTrader,
charts just pop up. And you think ìOh this is great,
I got my data and everything looks goodî, but the problem is that most of the time the dataís junk.
So, you canít actually rely on it and do any serious analysis. And getting the data and getting it formatted to MetaTrader
is the most convoluted, difficult problem that I face everyday, because this is the platform that we deal with
and everybody has problems with their back testing and getting familiar, so itís a pet peeve of mine.
Attendee: I didnít know that. Shaun Overton: Yeah, itís very bad data.
Attendee 01: Whatís the resolution for…? Attendee: Donít use it for real money. Thatís what I can tell you.
Shaun Overton: Yeah. It, itís good enough if you wanna trade every couple of minutes
and youíre not super execution sensitive and you are just trying to get something out cheaply.
Like… so, if you try every 4 hours and you trade 3 currencies, perfectly fine. Thereís, MetaTraderís fantastic for that.
But if you try to scout or day trade or trade 20 different currencies at the same time,
thatís a disaster, because this isnít multi-threaded so every time you push and order into the market,
it can only handle it one at a time. So if you have 5 orders going together,
this one has to finish, and you have all the latency in the middle
where they connect and then bounce and the trade confirm okay,
now you can do this guide to repeat the process. So if youíre pushing too many orders through,
this will, will choke. Attendee: Is that problem still a case with MetaTrader 5
or was that just with 4? Shaun Overton: Um… you know honestly, I donít know.
We havenít messed with MetaTrader 5 enough in live market to have that come up yet.
Attendee: Whatís the resolution of the data that you can download?
Shaun Overton: Itís very restricted, so MetaTrader doesnít offer very much in the way
of flexible timeframes and itís actually a big problem in the market.
So, with MetaTrader you get 1 minute, 5 minute, 15 minute, 30 minute and you get 1 hour, 4 hour,
daily, weekly, monthly. Attendee: And how much can you download at once?
Shaun Overton: Ah, it depends on the broker and your settings. So if you are looking at 1 minute data,
some of it will let you load a month and thatís it. Some of them have no throttle on it
and as long as you tell it to stop auto scrolling, which is… so when you load a chart, it, ah…
if you scroll this way and then a new tick comes in, it pushes you back to the present.
Attendee: Very sophisticated. Shaun Overton: Yeah…
Attendee: But I mean can you store the data… I mean, why not just write a routine then that
just collects everything that you need? Shaun Overton: You could do that.
Um, yeah, that would be fine, but for people that want to do like 5 years of testing,
thatís a long waiting period. Attendee: Even with continuous contracts thatís limited
by what the chart will hold, itís a, itís a big mess. Itís a horrible mess.
Shaun Overton: Yeah, you have to… the brokers donít want you to have all this stuff.
Itís not like a conspiracy. Itís the more people pinning their servers
and downloading 50mb worth of data times 10,000 users… not fun.
Ah, NinjaTrader is a lot different. They donít give you data,
but they make it easy to get data, but you have to go buy it. So, MetaTrader, you get data for free,
it loads very simple but itís junk. NinjaTrader, you have to find somebody to give you the data.
So, there are some paid options, there are some, thereís one thatís free called Connetic,
and they give you end of day data. And then with the brokerage, they have their own policies.
So, MetaTrader isnít some stainable platform, you just download MetaTrader,
youíre hooked up to a Broker. NinjaTrader is designed to be something totally different
where you have all these. So… Iíve totally run out of space. The problem when youíre programming all this stuff
and youíre trading in the live market, if you program to a broker specific platform,
you probably spend 4 to 6 months developing and testing it and getting it working,
and a lot of money and time. So that if you go to start trading live
and youíre not happy with the broker, too bad, you married them.
So, what NinjaTrader did is, letís say that this is Brokers A, B and C,
NinjaTrader is here. They shove the broker API on top of it
so that you can write your strategy here in NinjaTrader, and then theyíve done all the integration
with every broker partner they have. So itís a different way to handle the same problem.
MetaTrader just goes to these brokers and say ìYou should use our platform. You should use our platformî.
So if you developed in MetaTrader, you can go switch on limb. If you have NinjaTrader, you can go switch on limb.
But itís not, itís, you donít just have NinjaTrader, you have to download the broker platform
then you download NinjaTrader, then you get everything hooked up and make sure itís nice,
which is effective and it does work but itís certainly not user friendly.
Thereís a steep learning curve, uh, with getting this all set-up
to the point where you can actually download historical data and start trading with your broker.
Once you have the data set-up though, NinjaTrader is awesome.
I love doing the analysis in NinjaTrader so I, it took me weeks to get to the point
where I was motivated enough to go through this process and figure it out.
But once I did, I was really happy with it because I know the dataís good and the analysis is great
and it runs C Sharp, so, thatís what I happen to program in most of the time.
So, for me, itís just a great fit. TradeStation, Multicharts, these are,
these offer the same quality of analytics as NinjaTrader, but again, the problem with TradeStation is,
just like I said, where if you program for a specific platform, youíve married your broker.
If you program with TradeStation, you trade with TradeStation,
but you canít go trade with anybody else. So it is easier to develop in and it is easier to, uh,
to test and all that stuff, but if youíre ever unsatisfied with the broker
because they slip you or because they charge bad commissions
or whatever goes wrong with your trading, you donít have any alternative,
you gotta go with your program or repeat everything from scratch.
Attendee: Uh, could you, could you elaborate more on, on the, ah, the reasons why I might be upset with the broker?
Shaun Overton: Sure. Yeah. Thereís all sorts of reasons youíll be upset with a broker.
I used to work as a broker, so I can give you a long and distinguished list
of all the angry calls Iíve had for 2 years. Okay, number one, bad service.
You call in and you talk to somebody that treats you like youíre talking to somebody at Chase or Wells Fargo.
Theyíre somebody in the call center and they donít really care about it.
Number two, bad execution. This one is going to be the big one.
And a client just asked me about this today because he thought it might be a programming problem.
So heís trying to buy at zero Dollars, and letís say the current price is 1.3180.
Okay, every single time he trades, the broker recodes and it says
ìUh, I just got your order and now the price is 81.î So he has to pay more every single time he trades.
Or, letís say that… my point is, they show him one price,
he places his trade and then they just play some games, and say ìOh thatís not where you can tradeî,
youíre gonna have to accept this worse price and then they donít give him the option to do something else.
Um, they have poor banking relationships so itís entirely possible that youíre dealing in…
this is especially true in Forex. I mean, there are literally brokers
in every jurisdiction around the world, from hugely developed capitals markets,
like the U.S. and the U.K., and then you have guys in Belize and Panama.
So, you have all the options and theyíll probably take your account,
but the quality of the bank is not all that great. Thereís no regulation or oversight, literally.
So itís just you and the business, and you take their word for it.
Um, you can have bad latency, which is really related to the poor execution.
Um, you can just have, literally, software goes down, so it crashes.
Iíve had that plenty of times. Attendee: Every firm is the same place as the custodian
which is… Shaun Overton: Yeah, you have conflicts of interests for a,
kind of, they play these kinds of games which is, I guess, yeah, itís another form of bad execution, but,
same, same story different details. Attendee: So, within all these things, is it,
like the same as having like airline loyalty that some people will always have bad experiences with United,
and some people will always like American Airlines, or something like that.
Shaun Overton: Yeah. Yeah, thatís, thatís marginally the case, except with Brokers
it tends to be even more well-founded. Attendee: Okay.
Shaun Overton: So, like, they, if I have a bad experience with United,
itís because somebody gave me bad service. Like, somebody bound off to me one too many times.
If I get bad execution, that doesnít have anything to do
with me having a bad experience, it has, you just stole my money, which is a lot more emotional.
You get really worked up when, ok, youíve gone through this whole process right.
Youíve done all the analysis, youíve been psyched about trading
for probably a couple of months, itís all youíre thinking about, you loan your life account,
and then they start stealing from you. Youíre not a happy camper.
Attendee: So is this, does that happen across the board? Do people try to play those games?
Shaun Overton: Yeah, thatís the biggest issue with algorithmic trading right now.
Um, thereís a great book that I read a couple of months ago and itís completely in line with this group.
Itís called ìDark Poolsî and itís basically about how all the institutional algorithmic tradings steals your money,
fractions of a penny at a time. And thatís all true, basically institutionalized corruption,
where itís supposed to be a fair and transparent market where everybody places.
There, thereís supposed to be 2 prices, the Bid and the Ask.
But whatís happened is that thereís all of this different places where you can trade.
This is for stocks, so letís say that this is Nasdaq. There all these different pools where you can trade
and they all have their own bid and ask. So sometimes,
because you have all these different connections, thereís latency between them.
So you have different prices, which means that you can make money for free.
Sometimes the Buy price will be less than the Sell Price. Itís like buying a car for $10,000
and then you need to lease it to someone for 11. Of course youíre gonna do that
if you find a dealer stupid enough to do that. But, that, nobodyís dumb enough to do that in real life.
The only reason those kind of things happen is because of technological accidents.
So you have that kind of thing, and then because you have all this different pools,
they give segment in priority. So in a free and fair market,
youíre supposed to really just have three types of Order; a Limit Order, a Stock Order and a Market Order.
Thatís it. So this is gaining what you got right now.
This is, I want better, something better than whatís currently in market.
This is I want something worse thatís currently in market. So if the price goes up, then Iíll buy.
That kind of thing. But what happens with these Dark Pools is that
they just make up their own rules. So theyíve come up with all these different kinds of orders,
and if you do enough volume and you spend enough money, then you get to cut in line.
And through a very long concluded process, you, you get to improve your pricing and steal peopleís money
because they donít get to the same advantages that you do. Attendee: Pay to prioritize or get ways.
Shaun Overton: Yes. Which is, theoretically illegal, but apparently not in practice.
And then this is the, this is a good example of broker specific platform.
Uh, itís efficient if you wanna do something really simple. So, my favorite example with all my clients is,
uh, Moving Average Cross. If the fast moving average Clock
crosses above the slow moving average, and then the buyers sell.
This is okay for that, but it has some enormously frustrating limitations,
like, it doesnít have loops. So you canít do a four loop in a think script.
You canít do DLLs or anything. So if you want to do
the absolute most bare minimum simple strategy possible, then this is okay.
For anything remotely sophisticated, these are okay. And, my personal preferences are these two,
just because theyíre stable. I use NinjaTrader more often because of the broker options,
but in terms of research quality, these are pretty much all the same in my opinion.
Attendee: Should you say for Multicharts, they do allow you to choose your data feed?
Shaun Overton: Yes. Attendee: Theyíre a clone of TradeStation,
but they didnít go with that aspect of it. Shaun Overton: Yeah, they do have some differences. You can choose your own feed,
there are multiple brokers thatíll take them. So ah, the, my uh,
really, my only gripe with TradeStation is why am I gonna pay for something thatís free
everywhere else and I donít want to be married to one broker.
Attendee: Well, um, how much do use the charting in these packages?

0:26:54.000,0:26:55.000
Shaun Overton: Personal?
Attendee: Yeah. Shaun Overton: Never.
Attendee: Well okay. Is that, uh, I mean for some, I have to edit,
ëcoz I have a job where I have to do some charting. And so, I use, I,
I got my, my regular package errors from ThinkOrSwim. I use their trading for their strong optionalitites…
Shaun Overton: Right. Attendee: …and, uh, I can,
charting is just totally deficient. So Iíve been looking at other packages.
Iíd be very interested in hearing what you said about other packageís charting capabilities.
Shaun Overton: This oneís far away the best, at least compared to this one.
This one is, okay, there are some candlesticks on the chart
and I can draw some lines, but thatís about it. NinjaTrader you can do all sorts of GUI stuff with,
so if you want to custom program something to look how you want it,
NinjaTrader will let you do it. Attendee: Okay.
Shaun Overton: Uh, itís not part of the documentation, but you can figure it out.
Attendee: Let me tell you what Iím going to do. I need to create, ah, the portfolios, first Iím to create portfolios
that were applicated by 10 different portfolio managers. Shaun Overton: Okay.
Attendee: Then I had to compare them from, also from the same time scale.
Shaun Overton: Right. Attendee: Um, they, theyíre composed of different securities.
Shaun Overton: Okay. Attendee: Furthermore,
what each portfolio manager was recommending, within a period of time, change all the time.
Shaun Overton: Yes. Attendee: So I had to, um, I had to,
I had to first start my graph with a specific date… Shaun Overton: Right.
Attendee: …which I canít do in ThinkOrSwim. Shaun Overton: Correct.
Attendee: Ah, then I had to, um, change all the graphs so they were relative scales.
Shaun Overton: Okay. Attendee: ëCoz otherwise you canít compare IBM and in Cisco.
You need the percentage in between 82 different securities. So then, I also had to take the base date, it had to be
different depending on when they made the recommendation, that had to be one.
Ah, it could be October 19th of last year, it could be February the 14th of this year, or whatever.
Okay, all those things. Can NinjaTrader, um, do those things? Shaun Overton: I wouldnít use any of these
for any kind of portfolio stuff. So, this is, this is the big problem that seems to be coming up more.
Or, these are all functioning on a single chart. So if youíre trading Euro,
Euro against the U.S. Dollar on NinjaTrader, you literally, okay, hereís my Euro dollar chart,
put a strategy on, put it on, now itís trading Euro dollar. And then Iíll trade Pound dollars, theyíll do the same thing.
Which is fine in isolation, but if youíre trying to manage a portfolio,
getting this to talk to this… the platformís not really designed to do it.
You can jury grid it, but itís not gonna be, like, you, itís gonna, a) involve a lot of manual labor,
and b) youíre gonna have to come up with your own way to talk to…
a) know what has been loaded, and to make those prioritizing decisions about.
Okay, a) you have this much capital allocated to you, do what youíre gonna do.
And then do that for everything to that portfolio. I have a guy who does that.
He has 70 or 80 charts open and, itís not fun, but he saved a lot of money
by doing in NinjaTrader. So it does work, but it comes at a cost. If youíre, you know,
if youíre trying to do this and youíre running a fund, you really, you just build from scratch.
Attendee: Right. Okay. Thatís fair enough. Yeah. Shaun Overton: Everybodyís pretty quiet.
Attendee: Iím still thinking about how much money I lost. Shaun Overton: Er, how many people
have actually traded before in the live market? Okay. And how many of you with blown up accounts before?
Attendee: The rest wonít admit it. Shaun Overton: Yeah. How long have you been trading?
Attendee: I would say 9, 10 months. Shaun Overton: Thatís pretty good!
Yeah, most people blow up in 6 months, so, youíre good. Seriously, the fact that you havenít destroyed the account
is very encouraging. The, the, I think itís, I forget what the stat is.
I think itís 90 or 95% of accounts open that U.S. works brokerages are closed
within near 6 months or 10 or a full year. And if you, itís…
Attendee: Itís not something they say on their front-end when theyíre selling it, I would assume.
Shaun Overton: They just overlooked that. I donít know how that happened. They have… the, the CFTC mandates that,
the CFTC is like the SCC for Forex. And the CFTC mandates that all Forex brokers in the U.S.,
now thereís only 9 or 10 of them, so itís really not a long list, publish every quarter
the percentage of their clients that made money. And on a low end, consistently from quarter to quarter,
youíll see something like 40% and youíll see maybe something as high as 65%.
So, and then you add that quarter by quarter, profitability doesnít have anything to do with scale either.
It means their account is one penny more or higher than it was 6 months ago.
And what happens is, with brokers, they have two ways of making money.
And this is only for Forex. Attendee: When you say Forex, is that a, just currency or…
Shaun Overton: Yes, thatís exclusively currencies. Although thatís kinda taken on silver and gold as well.
Yeah, so, right, I have Forex equals currencies… and I probably shouldíve asked this in the beginning,
but, is everybody familiar with the term Forex? You know how it works?
Attendee: Exchange. Shaun Overton: Yes. Then you know why you can make money
when the currencies… okay. I just want to make sure when Iím talking Iím making sense.
Okay, two ways brokers can make money: Commissions or Risk.
Commissions are, just like in Stock Market, you buy stock, you pay a penny a share, thatís it.
For Exes, the same thing, ah, except the one thing thatís unique about it
is that theyíll mark up the spread. So, theyíll say that
theyíre receiving this bid and ask from the bank, and then theyíll make it wider.
So itís kind of like what I described to you before, about buying a $10,000 and selling it for 11.
They might show you, the customer… sorry, 91. They might show you 91.
And then, as soon as you take the trade, they immediately flip it and sell it to the bank for 90.
So thatís, thatís how they make their money. And thatís, itís, a mark up in the spread.
And thatís all disclosed. So they have to tell you that we make our money
by marking up the spread, itís not free. The second is way more popular.
And, frequently a warning sign of who youíre working with is if the broker makes his money by taking on trading list.
And the reason this is lucrative is that… you were talking about
brokers publishing their quarterly reports, and where, basically, 40-60%, appliance make money every month.
But what happens is, qua… the, us type traders, just regular guys off the street,
they make money when volatility is low. But when itís high, they get utterly decimated.
So, the reason that is, is when you talk to most traders, theyíll always talk about risk managing,
you gotta manage your risk, you gotta make sure that you live to fight another day.
But for most people thatís a bunch of hot air, and what they do is they… say Iím losing,
the price is going this way, they bought here, and then, theyíre like ìUh, itís coming back.
Uh, itís coming back. Uh, itís coming back. Uh…î that way. And then they donít have any money left.
So, their account, if youíre on the opposite side of that, is now the property of the broker.
But when volatility is low, peopleís natural tendency to do that,
works in their favour, temporarily. Because the price, quote and unquote, always comes back.
So, when volatility is low, and when I was a broker, it was 2005,
and in Forex, that was the absolute record of low volatility ever.
And so the firm I worked for was not making anywhere near as much money as it hoped.
But then in 2008, volatility was the opposite and then voila, itís raining money.
Because when volatility goes up, the brokers want those accounts.
And the reason is, not only do most… this is kind of, doesnít,
this doesnít make statistics stand out as much as it should. So, my favorite example is… when I was a broker,
I worked on funds called ìThe Sentiment Fundî, and all it did was give you a ratio of the number of people
that were long and the number of people that were short. Attendee: From what… from what source?
Shaun Overton: Well youíre a broker, so you have a giant book of a hundred thousand accounts.
Attendee: Your clients? Shaun Overton: Your clients.
Attendee: Client based… client based? Shaun Overton: Yeah.
Attendee: Okay. Shaun Overton: So you could say that 80% of clients are short,
the Dollar against the Yen. And then… er, oh sorry, 80% are long and 20% are short.
But, if this was the current market it would be the opposite ëcoz itís been skyrocketing.
Okay, why is this important? Because the retail crowd is always wrong.
Always. Like, itís a guarantee fail. Okay, so if you have access to the order book,
you know where the market is going. Itís just a matter of time,
because these guys are gonna get pressured. So, in this scenario that I highlighted…
okay, Iím frustrating myself coz I keep switching which example Iím using. Letís say that only 20% are long,
80% are short, in this example. The way this ratio works is it grows over time.
So here, it might be something like 50-50 where you have this nice flat market
and not a lotís happening. But as the market goes down,
these guys wonít just hang into the trade, theyíll start, theyíll buy here, and they theyíll buy here.
And then you reach this critical point, where, okay, Iím starting to run out of money, I…
I need to start closing some trades or Iím gonna go short. Then you have the train traders that are hopping in,
and theyíre saying ìWow this thing is starting to get some leg,
so Iím gonna go shortî. And all these buys start to come in and…
more selling pressure, so that you get those overextended moves.
As this is happening, the ratio starts getting worse. So here, you might start at 50.
And when you get to some extreme point, like here, it might be 20. And then some currencies, uh…
I mean, literally, can be 10%, are short, into this kind of move and the other 90%
are smart enough to know itís coming back. If youíre the broker
and you have access to this kind of information, itís a gold mine. And if youíre taking on the risk
and manage it appropriately, then you can make a lot of money.
But the problem is that, for a business, itís not really nice cash flow.
It all comes in and over the course of 2 months and the other 10 months,
theyíre just hemorrhaging money. Attendee: Got a question on that point.
Iíve noticed Fidelity will make that information available to the customers now,
as far as what people are trading, percentage, whatever, a variety of items like that.
Are you… do you see… is that common amongst brokers these days?
Theyíre becoming kind of a standard thing. Shaun Overton: Yeah, absolutely. So when I,
when I was doing this manage fund and they had a strategy built around it.
This is public information, so anybody that had an account there
could go see how this was and trade on it. Ah, and now itís even more public than it used to be.
And thereís a couple of different, uh, websites that are Forex specific,
where all of this stuff is published. And theyíre not brokers, which is the interesting thing.
So, MyFxBook is one of the ones that I like to use and… Attendee: OANDA.
Shaun Overton: OANDA. Yeah, OANDA publishes their data too.
For MyFxBook, itís everybody that has a MetaTrader account, which is probably, like 70% of retail Forex traders.
They hook up to this MyFxBook and it gives them all these cool analytics and stuff.
But theyíre reading the market positioning from all of these 100,000 customers all at the same time.
So they have something, like $100,000,000 of client funds in different positioning
that they get to see every minute that the marketís open. And you can just go on their website and look at it
and see minute by minute… oh this is… and they have graphs now, it goes back to…
and this is relatively new, it only goes back to, I think, October. You can see the positioning, minute by minute,
in all the major currency pairs, which I think is awesome.
Iíve long wanted to build a long strategy around it, I just have not have the time.
But itís a bunch of potential quantitative traders. Uh, that would be my place to look because
that strategy works… really well. Attendee: Just bet against the craft.
Shaun Overton: Thatís it. Ah, the ratio, yeah, the thing that we use in the fund was a 3:1 ratio.
So if anybody, if any group is 75% one direction, pretty good bet that if… and then you can add in other filters.
The best is if, if, if volatility is high and theyíre all betting one way, then…
you know, that far. Attendee: How, howís it look stock market wise right now?
Is it out of whack right now like that? Shaun Overton: My… personally, I wouldnít participate
because I donít know where the money is coming from. So, I just tell him, before the meeting started, that
most retail money has been flowing out of the stock market for…
I donít know, probably last 2-3 years. Pretty much since 2008,
where everybody took those massive losses and their pensions and their mutual funds,
and everybodyís just sick of it, and then pulling out money.
And except for a couple of months, for the most part itís just been 10s of billions of dollars
that are withdrawn from stock markets since the last crash. And then you have all the negative news, like flash crash.
And I donít think people know the details, but I do think that thereís a growing sense of
ìI shouldnít trust this and participate because I think this is not in my favorî.
Attendee: Appreciate it. Shaun Overton: I donít know if that really relates
to Forex platforms at all. This is sketchy, but hopefully youíre finding it interesting.
Attendee: What percentage of the market movement and the psychology of the market
is driven by institutional money as opposed to grandma gets upset and sells her (inaudible)? Shaun Overton: Okay, thereís a really big news story
that was on CNBC recently, involving a Carl Icahan and some other big fund manager.
I canít remember his name. Attendee: Bill Ackman.
Shaun Overton: Yeah, so itís Icahan and Ackman. They have this ongoing feud about the company in Herbal Life.
Attendee: Okay. Shaun Overton: Okay. Icahan swerves up and down.
This is a phenomenal company that has great prospects, and he thinks itís got a 50 or $60 target and is trading her at 40.
And this guy thinks itís a zero. Well, CNBC had some unusually un-entertaining television
where these guys are basically just yelling at each other for 15 minutes.
This saying ìThis is going upî and ìThis is going zeroî. Well, the, the interesting thing about this is that
the data happened… the, the market was actually down, but the shares that Icahan was buying Herbal Life
were going up pretty dramatically. So, whatís been happening
is that he went from a very nominal stake, like 2%, to way over 10%, where now,
heís had to disclose it as a major… ah, holding in his fund. So the point where it seems like itís just a big
one of these to Ackman, because itís literally the day that they got in this feud.
Heís messing with the stock price and buying, you know, 10% of the shares that are available everyday.
Attendee: Yeah, so Herbal Life is the… ah, thinly traded?
Shaun Overton: No, itís a big company. Yeah. But these guys also, you know, they manage billions of dollars.
So, if you wanna just have a, you know, a little fight with your favorite enemy, you could do that.
And if youíve got the money to do it, then you can move that stock around as much as you want.
As long as you got the money to do it. Attendee: So I… does that happen in the, in,
in the Forex market, futures market? It doesnít seem that youíd be able to do this sort of thing
to the expert. Shaun Overton: No. Ah, in Forex,
you would have to be an institution to possibly move the market,
barring some little exception, like… Iíve heard of colleagues that have, you know,
some wealthy guy in the Middle East with $100,000,000 to blow with it.
At least I thought a yard… a yard is a billion. Okay, I have… ah, you just… wants to gamble,
and for no apparent reason just buys too much, and then now he owes a billion Euros,
and he bought it in 5minutes. Thatís gonna affect the market.
Attendee: I see. Okay. Shaun Overton: But,
for the guys that are participating on retail level , the average account size is 10 or $20,000.
In Forex thatís, yeah, you donít know thatís happening. Attendee: Yeah. So the institutional guys canít move
the market quite as, as much in, in, in the futures and in the Forex market
as they possibly could in the equity market? Shaun Overton: Correct. Because…
and thatís just a function of equity. So if youíre, if youíre investing on a penny stock,
I can move a penny stock. Attendee: Yeah, yeah.
Shaun Overton: But, yeah, when you start going up to
0:46:06.000,0:46:08.000
small cabinet cap…
Attendee: Right. Shaun Overton: The higher the evaluation
and the higher the amount of money traded, the more people, messing with the prices.
Attendee: Yeah, last Christmas… not, not this Christmas, last Christmas,
a drunk trader bought 545,000,000 of bright crude, and market, sure enough, slowly but surely, went up…
Shaun Overton: Yup. Attendee: …and they finally traced it back to one dude.
Attendee 01: He was literally drunk? Attendee: So…
Shaun Overton: Yeah. You think it canít happen, but it really is just the most trivial reasons
with these guys with way too much money under management. But… you just,
you know, stupid things happen, and it really… Attendee: Whatís that one word that has
the picture of Tyler Durden and I canít remember. Shaun Overton: Thatís Zero Hedge.
Attendee: Zero Hedge, yeah. They got all sorts good articles about that.
Shaun Overton: I love that. Yeah. The… this group topic of algorithmic trading
is probably not popular on that site. The… yeah, I love Zero Hedge, I read it everyday.
And thatís actually where I got the Icahan Ackman feuds. They did a hilarious little piece,
and they used a much more colourful language than I did. But, uh, they, eh…
if you wanna know whatís happening macro, Zero Hedge is a great place.
Attendee: They were saying that they have a full quarter of a second now inside the exchange
before it has to leave the exchange. Shaun Overton: Really?
Attendee: Full quarter of a second. Shaun Overton: Okay. Interesting.
Attendee: Thatís CMEs. Shaun Overton: Okay.
Attendee: How are the events in the country affect the, um, the growth and the graphs?
Shaun Overton: Well, it depends. But for the most part, the driver of currencies are interest rates.
Thatís the theory. So, if you look at this from the non-speculation standpoint,
you just look at it from traditional trade back in the 1970s. You might look at a coun…
a country that has phenomenal exports and huge gro, growth opportunities. And… okay Iím looking at Mongolia,
and they have something like 15% interest rates. Attendee: Sorry.
Shaun Overton: And then here in the U.S. were making zero. So, what sounds more appealing?
So what most people do is they sell into the currency that has the higher interest rate.
Ah, and as these interest rates rise over time, people tend to follow and chase that rate.
So thatís generally what causes the currencies to move. But, weíre in the middle of a currency war,
where everybody wants their interest rate to be zero. So, itís totally skewed.
Because the way the market should quote unquote be is, really, everybody raising to keep their rates at zero,
and whoever gets their rate to zero, and most persuasively convinces everybody that
it will be at zero for the longest period of time, will have the weakest currency.
So, if you think about a country like Japan, which has been the hot topic in the currency market recently.
Theyíre an exporter, theyíre an island, theyíre totally dependent on their exports.
So when you look at a currency pair like Dollar Yen, which literally dating back to the 1970s,
looks like that. So, if youíre interested in ultra-long term trading, thatís…
the Yen, after World War 2, traded at something like 500 and now itís been at 90, and itís been as low as 70.
The problem with Japan is that when the prices go down, what that means is that the Dollar is weakening
and the Yen is going up. And if youíre an exporter, thatís a big problem.
Because… okay, letís… about 5 years ago, the Yen traded at 120.
Okay, so, now itís increased in value by a third in 5 years. So Toyota is selling cars for,
I donít know what appropriate price is in Yen… maybe 100,000 Yen?
Actually thatís probably way too low… in 500,000 Yen? Well, itís still 500,000 Yen,
but when you put that back in Dollar terms, itís probably going from 12,000 to 16,000.
Attendee: Right. Shaun Overton: So, for Toyota, they are getting crushed
when the Yen keeps gaining value. Attendee: Yeah.
Shaun Overton: And when you say thereís a currency war, what that means is that
all these people are trying to improve their exports by destroying the currencies.
So the best way to do that is to make it completely unattractive.
So, okay, you wanna put your money in Yen. Fine, but Iím not gonna pay you any interest.
Well, over time, if nobody else was playing this game, you would win. Yen will start going like that.
And uh, there have been a lot of hiccups along the way, where the Yen will go up
100 or 15 hundred pips or so. And a lot happens from the bank intervening in the market
and actively buying the currency. But for the most part,
if you want to destroy the currency, you can. The current environment is that, itís not just Japan,
now itís the U.S., now itís Switzerland, the Euro to a large extent.
South Korea is getting involved. You have all these, uh, yeah, you have all,
oh, yeah… then you have Great Britain which is also doing its own quantitative easing.
So pretty much every major currency in the world is trying to devaluate self-competitively,
where everybody wants to be zero, but everybody canít be zero.
So, instead of speculating on who is going to have the best interest rate in the next 6 months,
what youíre really speculating on is who is going to do the worst job at destroying a currency,
because theyíll be the best, which is insane. Attendee: How… howís the U.S. doing? Are we top?
Shaun Overton: We should be. I donít know why people think the Dollar is a great place.
I think itís a disaster. But uh, yeah, weíre… I think itís um… Attendee: Is it the best disaster?
Shaun Overton: So far itís been the best disaster, uh, depending on your perspective.
Uh, Japan has been most convincingly destroying its currency why… which is why the last 6 months itís been hanging up.
I think the Euro is a disaster in the long run in terms of functioning as a currency.
But the reason itís been so strong is that all of the dysfunction the Euro,
the same reason that they canít kick Greece out, that they canít kick Spain out,
that they canít make anybody follow their own rules. Also it prevents them from being good at,
then dealing interest rates. So theyíre still in the 1 point 1 in a quarter range.
Whereas the… news over last year, up until the last Feb within this week,
has been, weíre gonna print money into infinity, weíre gonna keep interest rates at zero for the next decade,
practically. So, here, everybodyís like ìYup, theyíre staying zero.
Nothing is gonna change.î And here, thereís no reason to think anythingís gonna change.
So, you see a little bit more movement into the Euro even though… I think, you know, it could collapse anytime.
But over the short run, looking at the interest rate, this is a better deal than that, for now.
Attendee: How about for the world currencies? That you… you introduce a component of social risk?
Shaun Overton: Yeah, and then you have other things. So, uh, obvious examples… September 11th, right?
The Dollarís going crazy, uh, social, war, those kind of things.
Uh, that changes it on the flip of a switch and everybodyís gonna start buying or selling,
depending on what the perceived risk is. Uh, I donít remember what happens when Fukushima
was going on in Japan, but I will wager that it was not good for the Yen.
Um, a lot of these other things arenít… like, you see the prices move in response to news,
but they donít move in any kind of fundamental way. Itís not like 911 happens and then…
Attendee: Yeah. Shaun Overton: Itís like, oh ok, yeah,
maybe I need to start rethinking this. I mean over the next 2 or 3 months,
this one really goes kaput. Uh, the way to see more is that,
when you have those big moves, those tend to be inflection points.
But if itís really gonna change peopleís permanent or outlook for the next 6 or 12 months,
it takes a while for the herd to change direction. Attendee: So, so third world countries,
where you have no… um, no long… no longer or the immediate term, uh,
guarantee that the currency is even remotely stable? Shaun Overton: Right.
Attendee: How do you trade those kinds of currencies? Shaun Overton: Well for the most part, you donít. Because…
Attendee: Stay out of those completely? Shaun Overton: Right. Because, I… I mean it depends,
if youíre doing algorithmic trading, you need liquidity. Because algorithmic trading…
Attendee: Yeah yeah yeah, youíre right. Shaun Overton: …because youíre, youíre pretty active.
Attendee: Right. Shaun Overton: So you need a thin spread.
If you trade something a little more exotic, like South African Rand…
Attendee: Yeah. Shaun Overton: …or Turkish Lira…
Attendee: So the spread goes up? Shaun Overton: It goes way up. Okay, this might be
8 or 9x as much as it cost to trade the Euro against the Dollar. Attendee: Got it.
Shaun Overton: So you can do this, and then when you go (inaudible) again
you start talking about trading, um, I donít know what to set an example… Chilean pesos.
Nobody is speculating in the Peso, so you can… there are, eh, there are algorithms,
and you can trade with HSBC. They specialize in that, that kind of thing, but…
your spreadís simply like that, so… Attendee: Yeah.
Shaun Overton: Youíre not going to be daytrading the peso. Well, not the Chilean peso, so maybe the Mexican Peso.
Attendee: I see. Okay. That makes sense. Attendee 01: Iíve been dying to ask you,
what do you think is the best strategy for somebody who has less than 50k in their account?
Shaun Overton: It… are you trading? Attendee 01: Trading futures.
Shaun Overton: Um… Attendee 01: Is… is it like a proportional system of some sort
or is it… is it, um, ëcoz Iím trying proportional systems and theyíve worked for a while.
But then, unfortunately, I was one of the fools who put NinjaTradersí 64-bit on this machine,
and had the SQL Server bug in it, where it hit the low of the day, it stayed at the low of the day.
Shaun Overton: Oh. Attendee 01: Yeah.
Shaun Overton: Thatís not good. I didnít know about that bug. Attendee 01: Yes. About 18 months later I found out.
Shaun Overton: Oh. Attendee 01: My neural network kept saying ìThis is great!î
One direction, it was great and I made money. Zero losses for 36 hours. And then slowly but surely,
the next 5 months a little bit, a little bit, everyday. Shaun Overton: Wow.
Attendee 01: But yeah, I just, it was… what is a good worthwhile strategy that you think,
for somebody like me, would be good for, I mean, you see peopleís strategies all day long?
Shaun Overton: Yeah, most peopleís strategies are random. So if youíre doing analysis kind of, like a, a pure quant,
the biggest problem is that the market is highly random.
Itís not entirely random, but itís really really random. So, my favorite example is,
I work with this professor who does financial markets, but he also works in cryptography.
And he did a little experiment where he decided to use market data
to see if that was random enough for him to generate in his cryptography.
He found it wasnít good enough. So itís, itís enough that it looks random,
but there are patterns in there that you can tease out. But doing that from, like, a data mining perspective,
is incredibly complex. So for somebody with our type size accounts,
I always recommend total opposite approach. Where instead of data mining and messing with MATLAB
and all those kind of things, itís just simple stuff like, ah,
like the strategy I mentioned, the, the mark positioning. I donít know what will… well, in futures,
itís a lot more long term, ëcoz you have the Commitments of Traders.
And that report will tell you on a weekly basis that X number of traders are long WTI,
and the commercials are doing the exact opposite of what the retail traders are.
And just like in a Forex market, the retail traders lose pretty consistently, not a guarantee.
And the commercials arenít always right. There are exceptions.
The most popular one was, ah, when gold started picking up in, I think, 2009.
Ah, that was one huge counter example, the commercials were short the whole way.
Ah, and they lost tons and tons and tons of money, the retail traders actually were right.
So itís not iron clad, but for the most part, when WTI went up to 140,
the retail traders are short the whole time. So, this oneís hard because it only is updated weekly.
So, it really depends on what and how actively youíre trying to be involved in the market.
Uh, but you, you hard press to get it ridiculously wrong. Uh, thatís probably a safer way.
For Forex, it would be some kind of sentiment so, I, whenever I, I try to trade algorithmically,
but whenever I just feel the need to pull a slot machine and start trading, I will use MyFxBook
and I just do a really simple strategy and… do you program? Attendee 01: Yeah, I, I built a neural network with a,
you know, a string based format, um, complex in processing. Shaun Overton: Okay.
Attendee 01: So it was in .NET, I set the first 64-bit. Unfortunately, NinjaTrader cannot.
But I was using proportional strategy where I took an institutional feed from ah, HotFX, the HotFX high,
and so I did proportional. So, the Euro has 12% of the market,
this other one has certain percentage of market, and then did a correlation on those,
to see what the proportion is in. The actual exchange traded half of that currency.
Shaun Overton: Okay. Attendee 01: And so, every, thereís change you can make,
you make a short change, like 8 currencies or 26 currencies or 100 or so.
And so, um, I guess where, Iím, Iím trying to find something opposites that work.
But something that you think is more reliable, because youíre talking about the sentiment analysis,
this sounds much more realistic. Shaun Overton: Yeah, it is, itís totally realistic.
Like they, these have, they have an API that you can use. And something Iíve been wanting to mess with for a while,
I just havenít gotten around to it, is looking at, um, long term. So, maybe, pull out the 4-Hour Chart,
and they have these cool lines where green is long and red is short and this is, obviously,
this is the bigger number and thatís the smaller number. So, whenever this line is above this line,
if, if these guys are short and this is long, well, there are more shorts than long, so I wanna be long.
And just try some then, I mean, itís really stupid simple. The lineís above this line, buy.
Attendee 01: Right. Shaun Overton: Um, and this, this Iíve only investigated by eye,
but it works. And because I know, I mean, they, they managed to make massive amounts of money doing
this type of strategy. And I eyeballed and it works fantastic.
It caught that Yen move, they went from 70 to 90.
Um, and itís, uh, itís still long, I think. Everybodyís still massively short, the Dollar against the Yen.
Attendee: Yeah, I see the 3 Ω to 4 typical window before like a, a currency will tend to, you know, retrace or invert.
But, um, doing this sounds much more, um, stable. Shaun Overton: Itís way more stable. The only thing is, uh,
and then you were mentioning, like the math behind it, is that the biggest obstacle for me with trading
is getting used to the distribution of returns. Because most people, when they think about trading,
is they want it to be an atm machine, where they go out everyday and they pull out a hundred bucks.
But the, the way the math in the market is, is it doesnít work that way.
So, like, 80% of the time itís just noise. Thereís nothing happening. And if you want you can play in it.
But the problem is that when these big tail events happen, you get massively destroyed.
So you can play this game, but youíre gonna give it all back when this happens.
And where you make the most money is when these kind of things happen
and the trends really start to blow up. This catches it and itís as simple as it gets.
I worked with a client that builds a very sophisticated neural network on this type of math,
and his returns are the exact same where itís 3 months of losing a percent, 2%, something like that.
And then you have one trade that gives you 25%. So, itís, you just have to,
and Iíve, Iíve messed with that a ton, where itís like, címon place a trade everyday.
You know,like, people want signals, they wanna see some action.
But it just, it doesnít work. So you gotta get rid of everything that you wanna see,
and just let the market do what itís gonna do, even if you donít quite like it.
And my opinion is that the best strategies donít make money consistently.
They just, you know, whenever itís there they grab it. But for the most part itís just…
Attendee: Ups and downs. Shaun Overton: Yeah, ups and downs.
Attendee: How much will you charge if I can show you my code? What, whatís your…
Shaun Overton: Iíll give you my card. Attendee: Benjamin here has a, ah,
he does, ah program called Quantopian and itís ah, itís created, it analyses, uh, market data and, whatís,
Iím not really sure if thatís the best way to describe it. But it does. Like NinjaTrader does,
where you can go and do back testing and find optimum trades…
Shaun Overton: Okay. Attendee: Does it, does it do any optimization?
Attendee 01: No. You, you, the idea of Quantopian is that you do everything yourself.
They just provide a platform for you to do it all. So itís more for the people that are interested
in doing their analysis themselves and not relying on the, I, I guess, the analysis and charts that seem to become
popular with MetaTrader and NinjaTrader. Shaun Overton: Yeah, itís handholding but a lot less of it.
Attendee: I have an in-house quant in Barclays now or… I have some in-house quant now,
they call it Standing Quant… Shaun Overton: Okay.
Attendee: …and Barclays is one of the big shops now. And Iím just trying to figure out the…
have you seen anybody use the, uh, uh, an adaptor in, ëcoz itís written in Python.
Anything for NinjaTrader and Python maybe, that we can pull in his…
all the stuff heís doing into analyzing our price stuff? Shaun Overton: Uh, when you start getting all these
platforms to talk together, it becomes a huge programming problem.
Because the way they organize everything is just completely different across different platforms. So…
Attendee: The reason why I ask is because NinjaTrader… I mean, uh, uh, TradeStation pays a lot,
it makes you pay a lot of money to do analysis. You, you know, buy a continuous contract,
and you just sack $300 for just one one CD for that one instrument.
Shaun Overton: Right. Attendee: But everything Benjaminís got going
seems like it should be a way we can short circuit the process. Attendee 01: But, I donít, I donít work for Quantum, again,
but they, theyíre trying to solve that problem. I mean, that, thatís what theyíre going to do.
Is… you can do everything in there, you can do all your analysis,
thatís what they want to do it in, I should say. Shaun Overton: Okay.
Attendee 01: Theyíre currently negotiating contracts with individual brokers.
They might try and go after MetaTrader and NinjaTrader as well too,
to solve some of these problems. Thereís reasonable expectations that you can do
language translations from, from Python to C Sharp. Itís not, itís not terrible and
you can do some Python design in that as well. Um, ëcoz there are…
Attendee: To save your data, is your data mostly in CSP file format?
Or what is, format is your data in usually? Attendee 01: Um, um, Quantopia.
Attendee: Ah. Or does it matter? Attendee 01: I think that, they will have,
the data is not, itís hidden. You, you canít really see what it is.
Attendee: Okay. Attendee 01: That, thatís how they,
they work out their contract with. Um, so you see, they have tick, tick data,
but you donít get access to it. You just, can test your,
you can do it with, write your back test solutions against it, but you canít see what the actual value does.
Shaun Overton: Ah, interesting. Attendee: Ah. Oh, okay.
Attendee 01: So you see, you get… Attendee: Like a Zipline,
I guess I got confused or something. I mustíve… Attendee 01: Yeah, no,
so Ziplineís our back tester and they can, you can get the, ah, the analysis is legitimate
but the values are not, um, representative of what the actual values are.
So, relative, something like that. Attendee: ëCoz we can put a neural network in NinjaTrader
and we can parse data with NinjaTrader from any feed. Thatís a good thing about NinjaTrader and itís,
itís only 49 bucks a month. So we can get it to play with Quantopian
and then have several instances of Quantopian running to parse that.
Attendee 01: Right. Well that, thatís how Quantopian is probably going to make money,
is they want to make that easy for you, but they want to, uh, thatís how…
Shaun Overton: Find a way to charge you for it? Attendee 01: Yeah, thatís how theyíre gonna charge you for it.
Attendee: Oh. Attendee 01: So you wanna take your,
your strategy and then connect it up to a broker to NinjaTrader. But for the moment, itís not, itís not possible there.
Itís a sort of limited Python environment, so, totally (inaudible).
Shaun Overton: Unfortunately. Attendee: So uh, yeah, so what you do most of the time
is you formalize peopleís trading strategies into automatic? Shaun Overton: Yes. So somebody will say…
a really simple strategy that I used earlier was the moving average.
Right, this one is the fast and this one is the slow, and the fast crosses above the slow.
I read these two moving averages and I see that this is above this,
but it was below before, so now we buy. Attendee: Right. Okay. So now, ah,
youíve been doing that for how long? Shaun Overton: 5 years.
Attendee: 5 years. So has there been a large increase in people wanting to formalize their…
Shaun Overton: Yeah! Attendee: …trading strategies?
Shaun Overton: Yeah, absolutely. Attendee: And do they trade them?
Shaun Overton: Yeah, itís kinda mixed. More of the market is people that want to buy
expert advisors in algorithms that other people have done. Attendee: Yeah.
Shaun Overton: Ah, but you didnít, there is a lot. I mean, my entire business is built around it.
Of people that have their own ideas and they want to automate, and then they tell us what it is
and then we make the software for them, usually in MetaTrader and NinjaTrader.
Attendee: Okay. So itís, itís a growing business. Shaun Overton: Yeah, definitely.
Attendee: I’ve always been amazed at these traders sitting at the screen all day doing something.
You ask him if they want to talk about what the strategy is, you realize itís totally automatable.
Shaun Overton: Yeah. Attendee: You should be walking the dog.
Shaun Overton: Well, thatís actually the most common. We have two kind of clients;
one are the guys who have been trading for a year or two, where they have experience
and they want to investigate ideas. Attendee: Yeah.
Shaun Overton: And then we have the guys that know what their strategy is,
theyíve been training for at least a decade, and some of them literally 25, 30 years.
And they just do this, I canít get up at 2 in the morning to trade the Euro anymore
and, okay, here are the rules, send it off and theyíre done. Attendee: Which is the most powerful, um,
Programming, you know, trading specific programming language?
Shaun Overton: I like NinjaTrader, ah, because I personally prefer C Sharp
and itís written in C Sharp and Iím… there are limitations. I, I have cranked NinjaTrader to the point where
it wonít run anymore. Um, so if youíre doing calculations
that you need to multithread, then you need to move it out of NinjaTrader.
But if itís getting, if itís somewhere between maxing out one processor
where you donít need to quite multithread, then NinjaTrader is the best solution.
MetaTrader will choke long before you get there. And TradeStation, if youíre gonna do that
youíre gonna have to write a DLL anyway. Attendee: Maybe we should, we should wrap up there.
If people want to stick around and socialize… Shaun Overton: Yeah.
Attendee: Letís thank uh… Shaun Overton: I hope you guys found it useful but,
Iíll be around so… Attendee: Yes.

Leave a Reply

Your email address will not be published. Required fields are marked *