If you find Quantblocks interesting, you should also look at Quantopian. (www.quantopian.com)
We're geared a bit more towards programmers. Rather than use blocks, our members develop their algorithms in Python. We have an in-browser IDE with a lot of smart auto-completion.
A few of our nifty features:
* free access to 10 years of by-minute historical data for all US stocks
* the writer of the algorithm owns the algorithm
* batteries included - all of your favorite Python math and science packages including Pandas and NumPy
* a robust backtester that models slippage, commissions, risk metrics, and more
We also have a community of quants and programmers who like talking about this kind of stuff. People share code, give advice, ask questions, etc.
Great work Dan, it is totally what I would be looking for when backtesting and strategy development. Do you mind sharing what js framework you used on the client side?
We wrote a bunch, and used a bunch, so there is no straight answer. A short list: highcharts for charting, jquery and underscore for the glue, crossfilter for data filtering, bootstrap for components, codemirror for the IDE, handlebars for templates, markdown for markdown, prettifier for code highlighting, and the list goes on.
Hi Yuri, not yet. We are looking at different possibilities for expanding the dataset, and the most popular right now is futures. Equity options do come up often, and are a close second for market data.
However, we are actually more excited about adding non-market data, because we want to bring more talent to 'algorithmic investment'. We hope our community can create algorithms that make buy/sell decisions based on more than just liquidity - fundamentals, reported data, qualitative news and research content. In other words, automating more of fundamental analysis and investment.
Looking at the terms & conditions, any algorithms that the users create become the intellectual property of Quantblocks. Is this just a tool to help users or is it a tool for the company to source the best algorithms?
The signup/confirm/login metaphor is really tired and annoying. Don't give me two screens of meaningless text, show me what your product does and let me interact with it before making me sign up. For example, open the strategy builder window up to anyone, but make me sign up before I can save my strategies.
On the product itself - security specific rules are fine but you're missing the point of backtesting and automated trading. The ability to spot patterns across any security (or perhaps in an industry) is the key here. I wish there were some way to specify this. I realize it's v1, but your rules are too simple for most traders.
Finally - you should call out that your market data is EOD. You should also offer VWAP (I know Xignite offers it) as a reasonable alternative for prices you "could have gotten" as a retail investor.
Also, the IP thing scares me, and I wouldn't use the service until you change it.
It is really too bad folks keep pushing technical trading platforms to the public. The promise of quick money to be made in a sea of billions of dollars might seem innocuous, but it's (a) bad for the market and (b) bad for society.
It's bad for the market because collectively, the buying and selling of shares based on anything other than company fundamentals (earnings, cash flow, projected growth, etc) distorts the price of the company. Unfortunately, because of our hunger to make money fast, there are too many of us (including large hedge-funds) playing this game, and the effects on price movement are very real. And when prices no longer reflect the company fundamentals, all sorts of bad things happen: Management is pressured to take extraordinary actions just to mitigate the market situation (stock splits/reverse-splits, stock buy-backs, accounting tricks, etc), employees freak out and quit, long-term investors get nervous, potentially fruitful M&As fail to happen, etc.
(The argument that technical/high-frequency trading improves the liquidity of the market is the biggest bullshit, cop-out answer ever: The only people that benefit from this type of instant liquidity are short-term, short-sighted traders ... such as the very people who advocate this type of trading, and not long-term investors!).
As for the societal cost, technical trading educates and perpetuates the myth that the stock market is a big gambling house, and not a means to become an owner of a company. To paraphrase Warren Buffett, every time you think about buying stock you should think of it the same way as if you were buying a mom-and-pop shop, like a pizza place. Is the price you're paying roughly equivalent to - or better than - how much you'd expect to make by pocketing the profits of the shop over the lifetime of the business?
I hate to leave completely negative feedback, so at least I'll give that this tool looks spiffy.
UPDATE: It seems like I'm getting a lot of responses from traders here. Already answered some, can't answer them all. Please sleep over these comments and think about what you could be doing with your time. Life's short.
Liquidity is the reason you can click a button on E-Trade and get your trades almost instantly, for between zero and a few more pennies than what traders would pay, plus a $9.99 commission. This was absolutely not the case before computer trading and electronic markets.
The "good old days" of human traders where when you'd pay fifty bucks plus one percent to a broker who would almost certainly front-run your trades, and the best spreads were at least 1/8. Nowdays the worst that could happen is an HFT might, perfectly legally, place an order a split second faster and extract fractions of a penny as they narrow the spread for everyone. Please forgive me if I don't consider this a disaster for investors, hedgers, mutual funds, and other consumers of markets that they can quickly execute orders close to the market price.
> Liquidity is the reason you can click a button on E-Trade and get your trades almost instantly
I personally don't care that my trade happens in a second or a minute because I'm a long-term investor - someone who wants to become a company owner, as I explained above, and not yet another hacker who couldn't care less about the long-term health of the company I'm buying.
My point is that if folks care about liquidity on the order of seconds/sub-seconds, they are themselves falling into the short-term trading BS (and consequently continuously distorting company prices).
Please stop perpetuating this insanity. Here's a timely article by Mark Cuban fresh out of the oven to help clarify my point:
Mark Cuban bemoans the good old days like any technical trader before they were beaten out by computers. He doesn't advocate trading on fundamentals. The battle cry against high frequency tradings seems to come primarily from individuals who were previously successful in specialist trading.
But yes, speak to us from the moral high ground. Cuban simply rode what essentially amounted to a pump-n-dump. He wasn't doing the promotion, but he sure as hell wasn't in it for the long term growth of the underlying company.
Here's a fantastic quote from Mr. Cuban: "What about fundamentals? Fundamentals is a word invented by sellers to find buyers.
Price-earnings ratios, price-sales, the present value of future cash flows, pick one. Fundamentals are merely metrics created to help stockbrokers sell stocks, and to give buyers reassurance when buying stocks. Even how profits are calculated is manipulated to give confidence to buyers."
Look, I'm talking about $50 and 1/8 spreads in 70s and 80s dollars. A low income investor from that time period might only have 3 or 4 digits to sock away in the market per year, Wall Street taking a huge bite every time. Then there are index funds and ETFs which need to get their costs as low as possible to benefit lay investors.
If you think big commissions and spreads are inconsequential, you are factually incorrect. If you think trading doesn't narrow spreads or lower costs, you are also incorrect. Finally, if you think trading qua trading distorts prices, there's almost no data that supports this and huge bodies of economic literature to the contrary. This isn't a matter of opinion.
>"I personally don't care that my trade happens in a second or a minute because I'm a long-term investor"
Long-term investor or not, you want liquidity to ensure you are getting in and getting out at the price you want. That's the big difference between investing in a liquid stock versus, say, your house.
The reason you can get in and out as a long-term investor is because of the liquidity provided by traders.
> I personally don't care that my trade happens in a second or a minute because I'm a long-term investor
The difference between a second and a minute could cost even the small investor many $thousands in the long term. I prefer milliseconds. If you're a buyer at $30 and a second is allowed, you're likely paying $30.30 or more due to front-running. Immediately after you buy it will be $30 again, so you lost 1+% in the delay.
The liquidity argument is not a "cop-out." Due to algotrading, spreads have been tightened and cost of entry/exit have been lowered. While there are some problems with algorithmic trading, the alternative (going back to the specialist model) is downright insane. If you think algotraders are "ruining" the market and society, then, the specialists and traders were actively scamming everyone.
I'm glad that you have at least done a little bit of research, unlike the majority of articles that came out after Knightmare, but algorithmic trading is a net positive for both the market and society.
Who gives a rats ass about shaving off spread points? Only "traders" - not long-term investors.
Also, I have done more than "a bit of research" on the topic (I'm ashamed I was once so innocent as to make money with stat arbitrage, now I know better).
> but algorithmic trading is a net positive for both the market and society
>Management is pressured to take extraordinary actions just to mitigate the market situation (stock splits/reverse-splits, stock buy-backs, accounting tricks, etc), employees freak out and quit, long-term investors get nervous, potentially fruitful M&As fail to happen, etc.
Oh please, if anything accounting tricks are more likely to be prevalent in a fundamentally traded market. When your company is traded solely upon earnings, you don't think you'd have more motivation to toy with those earnings numbers?
>As for the societal cost, technical trading educates and perpetuates the myth that the stock market is a big gambling house, and not a means to become an owner of a company. To paraphrase Warren Buffett, every time you think about buying stock you should think of it the same way as if you were buying a mom-and-pop shop, like a pizza place. Is the price you're paying roughly equivalent to - or better than - how much you'd expect to make by pocketing the profits of the shop over the lifetime of the business?
What is this? Some kind of argument from the moral high ground? The people who treat the stock market as a gambling opportunity will quickly be liquidated, and I can't imagine why one should feel sorry for them.
> It is really too bad folks keep pushing technical trading platforms to the public. The promise of quick money to be made in a sea of billions of dollars might seem innocuous, but it's (a) bad for the market and (b) bad for society.
Whoa. Get off the high horse. I don't disagree with you, but the problem is not with "the folks who keep pushing stuff to the public".
It's no different than a lottery, a casino, or selling cigarettes -- or for that matter, sugared drinks or unhealthy french fries.
Many activities are self-harming and potentially society harming. Usually, as a "free" society, we find the self harm acceptable, and the society harm acceptable to a point (based on magnitude of effect to entire society), mostly because experimentation IS required.
I read your post as equivalent to "I hope people who make and sell cigarettes would stop, because it's not good for anyone other than themselves". Which I agree with, but it's a useless rant.
The only way to fight this is regulatory. However, thanks to regulatory capture, a significant revision will happen, if at all, only after the next huge crisis.
There are many forms of algorithmic trading--some harmful (including playing games with the order book, canceling orders, etc). The types that this site seem to assist with are not those types.
Pushing technical trading platforms to the public? I don't follow how there is an overgeneralized public here, or how they are pushing something. They are releasing a tool to those who want to use it.
What companies like this do makes the RESEARCH for quant trading available to those who otherwise wouldn't have access. The organizations and people with the means to buy or build their own tools to do this were already doing it. Whether or not they are doing something that benefits the greater good is neither here nor there.
Saying those tools shouldn't be available to the general public is analogous to the old gun ownership argument. If guns are outlawed, only the outlaws will have guns. If only a limited few have the means to do quant research, it puts those who don't have access to those tools at a dangerous disadvantage.
Yes, there are risks to any sort of paradigm shift, but limiting the accessibility will hurt those who don't have access. If you (being someone who doesn't have the means) are unable to adapt to the changes, then you perish.
I doubt you or I are actually qualified to comment on whether the net externalities of HFT are positive or negative. You are incorrect that there are no positive externalities. You might not care about your trades happening quickly, but that liquidity is what allows them to execute at a good price and also a part of what makes indexing possible.
I totally agree on the wisdom of unsophisticated investors who think they can back test some "strategy" based on trends and support levels and so on and then actually go and make money. It's unlikely to work, they probably have no understanding of what their actual risks are (I'm talking probabilities, not just value at risk at any given time), and they are losing money on trading fees unless they can actually point to some competitive advantage they have over other traders.
I agree with a lot of what you say, but I think it is important to point out that algorithms can invest too. They don't only trade. Anytime you use software to buy or sell securities, people assume it is technical trading, or worse, high-frequency trading. The truth is there aren't enough people working on making algorithms that invest rather than trade, because the ROI for trading is more apparent.
I work at Quantopian, and our goal is to make it possible for more people to explore algorithmic investing. I agree the social utility of from increasing liquidity from current levels is at best diminishing returns, but algorithms could bring the same drop in costs and increase in quality for money management that it brought to trading. That's a benefit to individuals, pension funds, charitable endowments, and anyone else that has to save and plan for the future. Financial professions couldn't have a worse rap these days, much of it deserved, but there are real social problems that require financial solutions. People need to save for retirement, plan for the kids' college tuition, and take on mortgages.
I think it is really good for society to have smart people work on investment management. Especially if they are automating, collaborating, and discussing their work openly - the opposite of today's Wall Street. I like that QuantBlock did something original, and I love that they are striving for really broad access.
We think the key to advancing algorithmic investment is to create more access so smart hackers can tinker with investment strategies. Those folks should be able to explore and test ideas/algos without spending a few years building a backtester, or a few years' of salary on data. That's why our backtester is free to use at quantopian.com, and why the source code will be released at PyData NYC (http://blog.quantopian.com/pydatanyc-here-we-come/).
A cautionary note on backtesting (ie, assessing how trading strategies would have performed over a historical period in time).
If an a posteriori probability distribution is a good fit for historical events, it doesn't mean in any way it is going to fit future data points. It may or may not.
There are more, statistically sounder methods that could be used to search the space of possible strategies, a la bootstrapping. These could be used to make more meaningful statements about the strategy, if the assumption is that previous statistics predict future statistics, which of course might also be false (black swans, etc.)
Given the vast amount of parameters that constitute a strategy, naive optimizing of the outcome of the strategy in the past is bound to produce overfitting in the majority of cases...
Actually, this is not accurate. Before relying on our models, we get also the opportunity to validate them by checking at what extent they successfully predicts future data points.
Simply for the next t-time periods let the model give the user an estimate of the future prices/rates, together with an estimate of how accurate the model expects these predictions to be. This would allow the user to build confidence in the algorithm strategy she/he came up with, before employing it on the open market.
I mean it in the naive sense - we can observe the past but not the future. It definitely makes sense to paper trade a strategy and see its predictiveness in real time.
Absolutely true, backtesting is not a guarantee of anything. That being said, I can't remember who said that "No model is perfect but some models are useful."
this is great. love the design, and I can totally see a broker acquiring this so their clients will make more trades on their backtested-to-be-profitable strategies!
Hey! This is very cool, very fast. One thing I would suggest is the ability to specify a list of securities rather than a single security. In my experience, if you're going to test a general heuristic for trading, you're going to want to test it across as broad of a spectrum as possible to determine if it really has some predictive abilities. Obviously, this requires a lot more computational power if you're applying your rules to everything in the Russell 3000 rather than just a single security. The results would then be the composite of all trades across the securities.
This is kind of neat. It would be even better if it let you work with the returns of each stock, rather than just with the price. In my experience most technical strategies work better, and are more stable, when they work with data in return space rather than price space.
Thanks for the feedback, that's definitely something we've been looking at. This is our initial feature-set that we're using to collect usage-statistics. If you have any other feedback, please feel free to let us know.
This is fantastic. Love the clean, straightforward interface. However I have a feeling that anyone knowledgeable enough to profit from this is probably already heavily invested in another system for doing these calculations and/or feel the rules system isn't flexible enough. One thing you could do is to implement sharing. ie, have a list of "top performing strategies measured from date of creation". Then let users subscribe to other users' "proven" strategies for a fee - you take half and the author takes half - this could be an alternative way to monetize the site as well attract the kind of people who make a living doing this - ie the kind of people you want talking about your product.
Possibly. If you integrate with a trading platform you could always let other users subscribe to a strategy without knowing what the strategy is (eg they only know it's gained X% vs the S&P 500 in the last few months, the author's profile and that 5000 other users have 5MM invested in this strategy). However I have a feeling that a) the money from commissions and b) the feeling of being at the top of a list will compel a lot of people to share their strategies - people blog about their stock picks constantly. Also I'd love to see some left-field data about these companies eg: Google search volume, Wikipedia edits/day, mentions on Twitter, press releases, sentiment analysis of news articles etc...
It's funny that you mention it because, as Rob said, we've been playing with the idea of adding a competitive edge, but the commissions is a cool idea. We were also considering a kind of achievements approach as well as a leaderboard but want to try and keep the idea of it as a tool as opposed to a toy.
Love the data idea, in fact surplus of data is already a bit of an issue; we're running out of screen real-estate.
of course - the alternative is to not share anything. With the data you have on your users, once you can isolate users who consistently come up with strategies that consistently make money after their creation date - well, you might as well set up a hedge fund on the side and give the product away for free.
Your data is going to be insanely valuable if you can get good analysts working on your platform.
I was wondering if you could elaborate as to why they feel this way? I would love to know that the top X % of traders are willing to play with my money in a transparent manner.
haha fair enough - thinking about it there's all kinds of nefarious ways you could use the data. Looking at your product a bit more, this might be complicating things a little and it's going to hammer your database but I'd love the ability to fit variables to a strategy - eg I want a strategy like "IF X day moving sample average for GOOG is better than Y day moving sample average for GOOG buy ELSE sell" But I don't really know the best X and Y, so I'd like to leave those as variables and have the software find the optimum values based on historical data.
we've all been salivating about the potential for an 'Optimise' button. It would be really cool, it has a few wrinkles though, e.g. there might not be a unique maximum solution, so how do you choose the best one? Once there are a few variable involved it may be a job for a genetic algo, but it would have to be in the background. All food for thought ;-)
Here's how I'd do it:
Don't let the variable be the stock itself because otherwise your search space will be too big... limit the number of stocks you can optimise on and constrict it to variables like "moving average duration". That way you can load the stock data onto the client side and do the processing there (presumably you're already doing this for your graphs anyway) - kick it off with a few generations of pretty loose genetic algorithm then when you get near your time limit switch to simulated annealing with the best handful of results to narrow down the answer as fast as possible. Keep it quick, say 2 seconds tops. Then what you can do is give the client a confidence rating so you can say something like "Found the following result with 89% confidence. Click here to perform a longer search for higher accuracy". Work out your confidence rating based on how bumpy the ride was.
You could keep the strategy hidden but allow users to follow the output signals. There are some sites which already do this or similar for both fundamental and technical strategies. Motley Fool comes to mind: http://caps.fool.com/
Do you guys have plans to add support for Forex trading too?
And, even if one of the strong points is "no coding required". Do you foresee at some point to let advanced users write trading strategies in some type of programming language?
Currently it's daily. One of the possible future extensions is into intra-day but it brings a whole load of extra technical problems. Honestly the problems look really fun, things like keeping the backtest running relatively fast, caching the large amounts of extra data and the added complexity of intra-day strategies. However at the moment we need more data on how people want to use it as that's a pretty big outlay on development time.
1) The leading x days from the moving averages on the default strategy are charted at 0 when they should be null and not visible on the chart.
2) If I mouse over the 20 day moving average block and change it to a sell block, then click run, my charts disappear into loading bars and I get a warning message up. However, there is no indication to an uninformed user what they've done wrong, why it's wrong, or how to fix it. This could use a bit of work.
This is very similar to an iPad app I've been developing - I like it a lot, you've done a good job!
1) Just pushed an update to fix that, thanks for pointing it out.
2) Syntax checking has been a big discussion point for us, we would love to be able to limit people to only valid blocks but also didn't want to have a giant block of validation javascript to deal with the drag and drop on the blocks. As for error messages, you're totally right; they're massively lacking. Difficulty is that we keep changing the way we're storing rules as we learn more about what users want to do, given that we didn't want to have to build a proper syntax parser until we had a more stable concept of how they'd be structured.
Out of interest, what sort of error message would be useful? As in, how technical e.g. "I don't understand how to if moving_avg > buy" or "There is an error at 'if moving_avg > buy' returns: Nil expected: boolean"
My app is slightly different and uses a patch panel style of interface rather than linear blocks. Every time the system is changed, the layout is parsed into code and any invalid paths are coloured red to show
a) where the problem originates
b) which blocks are affected by the problem
Additionally there is a crude type system that prevents the wrong type of signal being passed to a block - e.g. a buy/sell datatype cannot be passed into a block which expects numerical time series data. I would say that you should prevent users from entering rules which make no sense as far as is possible..
I guess it depends how technical your users are. I'm attempting to stay away from terms such as nil or boolean, so I would go for something like your first example, but with more information, perhaps like:
"I don't know how to follow the rule "if moving_avg > buy" because xyz" which sounds friendly and non-technical but hopefully xyz will help them understand why it doesn't work and learn from their mistake.
I built something similar for my own personal use, so it's very interesting to see something similar polished for mass consumption.
The only thing I'm confused about is what you get in the Pro version..
"3 Months of Unlimited Access to Every Area of QuantBlocks"
How is the user supposed to know what "every area" is? How could anyone expect to know what they get with a subscription? Most importantly, how could they know if a particular symbol would be available in the pro version, what about futures, forex, etc?
So what if we said "3 Months of Unlimited Access to Every Daily Upgrade" ??
We're doing daily agile upgrades so we're trying to communicate that they'll get access not only to ALL of the global equities data, but also all of the updates that we make in terms of ability to save strategies, indicators, etc.
It might just be me, but I find it confusing that the upgrade statement says "a one-time payment of $39" and then the first checkbox says "3 months of access".
Does this mean I pay $39 every quarter to continue access? In which case why is it a one-time payment?
Hacker news has throttled Rob's replying (too fast), he says:
Thanks for the great feedback. We've just pushed an update that tries to clarify by removing "One time payment" and removing the "every area of Quantblocks" language.
If you really want quant stuff, then use R, Octave||Matlab or even Excel. MA, RSI test are available on Internet and works with all browser.
By the way - it does work well with my Chrome and you should move a message about supported browsers on the first page or before I fill registration form, not after.
I like the idea of the site, but only having 5 stocks available to test against doesn't allow me to get comfortable enough with the platform to pony up for a paid account.
Additionally, the choices you've made for inclusion could be more diverse. Where is SPY, SQQQ, GLD, etc?
I'd really like to be able to model the effects of rebalancing to a static or tactical asset allocation, but it seems that the only strategies that matter to the technical camp is thresholds based on trailing n days moving average.
Great job! Everything is really smooth. As a fellow developer who's interested in finance, I would love to discuss (if you don't mind) about your approach and effort.
That's useful to know. Is that daily data? Have you found it to be reliable? When I started testing moving-average based strategies, my blog post about it ended up being largely about data issues, from both Google and Yahoo finance: http://grahamstratton.org/straightornamental/entries/movinga...
An extreme example was an opening value off by a factor of 100 on one day.
In my experience, even historical or realtime data from Reuters/Bloomberg/CQG can have serious errors such as that - every trading system should be sanity checking its market data!!
Super interesting! Could you shed some light on to how much you are spending on the data? Is it in the hundreds, thousands or ten-thousands? Is it feasible for one person to bootstrap a website with Xignite?
I've had a lot of problems with Yahoo not having complete historical data, especially on thinly traded or OTC stocks. I wouldn't recommend using their data in a production application.
yeah that caught us out - we found that even though the data is freely available from those sources, you're not allowed to include that data in a commercial app :-(
Oh really? That's interesting - does that apply even if it's not distributed with the commercial app? I'm sure I've used trading platforms before which use Yahoo...
Our understanding of the Yahoo/google T&Cs was that we couldn't aim to profit from their data (http://finance.yahoo.com/badges/tos). We may be being overcautious but we didn't want to risk getting into trouble ;-)
Xignite allows us to show the full historical data of an instrument, but because we're taking a lean approach at testing how people use it, we're limiting them to a few years.
So judging from the the screenshots... Buy Apple? Yea that would of totally worked. Now for any other stocks, probably not. Also you can see it's less than the simple buy and hold.
Seriously I don't know why people even try trading in their free time.
Now, that being said, I'll take a bet against anyone who thinks they can find a profitable trading strategy using this product :D
Relax. The tool lets you create a strategy and back-test it. It is selling the fact that you can do this without coding, it is not trying to sell you 'Long - AAPL' strategy.
We're geared a bit more towards programmers. Rather than use blocks, our members develop their algorithms in Python. We have an in-browser IDE with a lot of smart auto-completion.
A few of our nifty features: * free access to 10 years of by-minute historical data for all US stocks * the writer of the algorithm owns the algorithm * batteries included - all of your favorite Python math and science packages including Pandas and NumPy * a robust backtester that models slippage, commissions, risk metrics, and more
We also have a community of quants and programmers who like talking about this kind of stuff. People share code, give advice, ask questions, etc.
Full disclosure: I work for Quantopian!
Happy hacking,
Dan Dunn