Back to Community
Important News for Our Community

Sit down before reading: Quantopian will be bringing you international equity support and a massive data catalog from one of the world’s largest data providers. We have partnered with FactSet to create unprecedented data access for our community.

We believe you can be more successful with more markets and more data. By partnering with FactSet, we are doubling down on our commitment to crowdsourcing and making allocations to the community. We will be integrating new data sources from FactSet and they will be freely available to the Quantopian Community. You'll be able to use the new data to research and develop strategies, enter the contest, pursue fund allocations, and collaborate with the rest of the community.

For years, we have been inundated with requests for enterprise sales and support from investment management firms large and small. We’ve always declined because we didn’t have a global team that could scale sales and support to make a successful SaaS business. FactSet has that scale today, and as part of our partnership will also begin selling a new service to their clients: Quantopian Enterprise.

We’ll be referring to the free service you know and love as Quantopian Community, and the new paid service as Quantopian Enterprise.

Quantopian Community will continue to have free access to all the functionality, new features, and existing data sources. Through our partnership with FactSet you will also begin to have free access to new data sources as we integrate, including: international market coverage, Supply Chain Relationships, RBICS (Revere Business Industry Classification System), Consensus Estimates, Ownership, and Geographic Revenue Exposure.

Quantopian Enterprise will run in parallel with our existing asset management business. The opportunity and path towards an allocation remains the same. Further, the community remains Quantopian’s north star for our product and for our company.

We are committed to maintaining parity between the Quantopian Community and Quantopian Enterprise platform capabilities. You will always have access to our latest and greatest quant workflow. There will be two important distinctions between Community and Enterprise:
Enterprise clients' algorithms will not be run through simulation or evaluated by Quantopian.
Quantopian Community access to the new FactSet data will be subject to a delay. For example, if a data source has a 1 year delay, Quantopian Community can access data through this date last year. All existing data sources will behave exactly as they do today.

When we started Quantopian, we wanted to bring quantitative finance knowledge, platform technology, and data to the world. Our dream was for anyone with the talent to have the opportunity to compete on a level playing field with industry titans. Partnering with FactSet brings us a big step closer to realizing that dream.

To learn more about this new initiative, visit: www.quantopian.com/factset.

Disclaimer

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by Quantopian. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. No information contained herein should be regarded as a suggestion to engage in or refrain from any investment-related course of action as none of Quantopian nor any of its affiliates is undertaking to provide investment advice, act as an adviser to any plan or entity subject to the Employee Retirement Income Security Act of 1974, as amended, individual retirement account or individual retirement annuity, or give advice in a fiduciary capacity with respect to the materials presented herein. If you are an individual retirement or other investor, contact your financial advisor or other fiduciary unrelated to Quantopian about whether any given investment idea, strategy, product or service described herein may be appropriate for your circumstances. All investments involve risk, including loss of principal. Quantopian makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances.

55 responses

Just out of curiosity, will Quantopian Enterprise include direct connection to the client's broker and automated execution of orders submitted by an algorithm? What about risk management ?

John,

In all of the years Quantopian has been running, I believe this is the greatest achievement so far. Many congratulations to all that have been working towards securing this partnership.

I used FactSet a while back whilst working in IB and also for my dissertation research whilst at University - I can say that it is without a doubt one of the best (if not the best!) data providers out there. I can't wait to include some of the new indicators in my algorithm, and expand my algorithm to trade international equities (I have long been requesting this feature, being based in the UK!)

Best of luck for the next steps,

Will

Hi Tim - Quantopian Enterprise includes the research and development workflow only, just like Quantopian Community. Our multi-factor risk model is available on Community and Enterprise.

Thanks for the explanation John, and lots of success with your new product!

Sounds expensive!

Congratulations Q, and all the best in your new business relationship.

As a user of Quantopian I think it is lot more easier to extend strategies that work on US Markets to other markets (Europe/Australia/Japan etc) than it is to write new strategies in say Futures or Forex. So thanks for expanding to new markets.

AFAIK the US dollar is hardcoded in Zipline and international markets require multiple currencies or at least, are traded in different currencies Two questions pop up then:

  1. Will the risk model be extended with FOrex risk? and
  2. Can we decide to define another currency as the base currency to eliminate the currency risk?

Great news!

Any live trading support...?

:(

Hi Fawce! I’m very happy about this. Getting more professional users of the platform, and international data sets, will really widen the user base and make my/our time investment go further.

... how does the zipline, hosted or open source, fit into this? Is the idea that institutional users deploy this for execution, or will they for sure have their own existing solution?

Hi Dan,
Just as zipline is used under the hood for the backtest capabilities on the community product, enterprise customers will use the same backtesting capability with zipline. Both products that we are providing are using the same shared platform. We are not providing any kind of execution services to Enterprise clients.

Thanks
Josh

Disclaimer

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by Quantopian. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. No information contained herein should be regarded as a suggestion to engage in or refrain from any investment-related course of action as none of Quantopian nor any of its affiliates is undertaking to provide investment advice, act as an adviser to any plan or entity subject to the Employee Retirement Income Security Act of 1974, as amended, individual retirement account or individual retirement annuity, or give advice in a fiduciary capacity with respect to the materials presented herein. If you are an individual retirement or other investor, contact your financial advisor or other fiduciary unrelated to Quantopian about whether any given investment idea, strategy, product or service described herein may be appropriate for your circumstances. All investments involve risk, including loss of principal. Quantopian makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances.

*"Quantopian Enterprise will run in parallel with our existing asset management business. The opportunity and path towards an allocation remains the same."
...

"Enterprise clients' algorithms will not be run through simulation or evaluated by Quantopian."*

Question: How will QE users be eligible for an allocation if QE algos are not subject to Q's simulation and evaluation? Would QE algos need to be submitted to the contest in order to 'give permission' for Q to evaluate and thereby be eligible for an allocation? Or is the QE offering targeted to clients that aren't necessarily interested in capital allocation from Q (e.g. firms rather than individuals)?

Thanks Fawce -

This all sounds good. I'm glad that you continue to move the platform forward on multiple fronts.

I don't know how to interpret these statements:

We will be integrating new data sources from FactSet and they will be freely available to the Quantopian Community. You'll be able to use the new data to research and develop strategies, enter the contest, pursue fund allocations, and collaborate with the rest of the community.

Quantopian Community access to the new FactSet data will be subject to a delay. For example, if a data source has a 1 year delay, Quantopian Community can access data through this date last year.

So I'm confused if the new data will be offered freely, or as a premium data set, with a fee required to obtain up-to-date data? Or perhaps there will be a mix of free and premium data available via FactSet?

Also, my understanding is that unless a fee is paid, premium data sets cannot be used in the contest, but could still be used to get an allocation (although in your example of a 1-year lag, it might not make sense to develop an algo, not knowing its performance for the last year).

Overall, please clarify, based on your definitions of "free" and "premium" on https://www.quantopian.com/data how the FactSet data will be offered? It sounds like in practice, for the contest and to do practical algo development for Q fund allocations, one would need to purchase the premium data, correct? Or is there a path within Quantopian Community to get access to the up-to-date data without a subscription?

@Joakim, yes, this is targeted towards firms and not individuals.

@Grant, think about access to the FactSet data in a new fashion. We don't expect the community to pay for access to FactSet data. However, in order to provide this data for free, there will always be a trailing hold out as Fawce describes above. Unlike our current premium data sets, FactSet data will be usable in contest algorithms without a subscription. We expect to provide more information on the mechanics of access to the data in the coming months.

Thanks!
Josh

Thanks Josh -

Presumably your present configuration of free/premium data sets will remain, and you'll add FactSet as another option? Or will there be changes to the existing free/premium data offerings, as well?

Sounds good that FactSet data will be usable in the contest algos without a subscription.

Thanks Josh. That's really great that we can use the FactSet data for free (with delay which is fair enough), including 'live' in the contest for true OOS. Very exciting!

@Josh, would you be able to share which international equities markets you are looking to roll out first (and any rough timeline)? Or will it be a 'big bang' roll out of all equities markets available from FactSet? Also, how will currency exposure be dealt with?

Great question. We have not yet finalized our implementation plans (we have been hard at work doing the research) but one thing I can absolutely guarantee - it won't be big bang. We will incrementally introduce markets. We will likely introduce a single incremental market and work from there. We have tentatively identified that first market but it would be a little premature to publicly commit.

Thanks Josh - appreciate the update. Exciting times to be a Quantopian!

Hi Josh -

I'm just starting to get my head around this announcement, and I'm wondering about your statement:

We are not providing any kind of execution services to Enterprise clients.

Wouldn't trading, using the Q platform in some fashion, be the end goal for some clients? I understand that Quantopian wouldn't provide the execution services, but wouldn't it make sense for there to be some path? It is a pretty big investment to do research and development in one API, and then to need to port it over to another one. If Q won't be providing execution services to institutional clients using Quantopian Enterprise, then who will?

This sounds like fantastic news. However Tim was dead on with what we all truly want. So, unfortunately, without the ability to execute our strategies for live trading, though I may invest in this to see what it offers, you are still very much out of sync with your user base. I think that if you're going to offer a paid product like this and prefix it with the word "Enterprise", live trading should absolutely be a feature. As I said before, I'd be first in line to to pay for it. Also, I'm not sure what the resistance is to allowing paying customers the capacity to live trade. It seems almost ridiculous, given the revenue potential. If your concern was that too many people would be utilizing the platform, creating chaos, then jack the price up, $200/month for live trading should eliminate all those individuals. Even if that left you with 10,000 serious users, that would be $24,000,000/year. Unless someone actually hit John over the head at a young age, with an actual faucet, this makes no sense to me at all.

Also, I agree 100% with Grant. No matter what you're selling here, ultimately we have to port it, which is what I'm working on coding now, as a result of these restrictions. The life blood of every business is recurring revenue. Once people utilize the platform and identify effective strategies, the moment they have to port the code elsewhere, this site becomes less relevant, and you will ultimately lose that customer. If you charge the customer to use your platform and trade here, you keep that customer, and the "recurring" life blood revenue that goes along with it.

Institutional clients have their own order management and compliance systems. Then we add a layer of SWIFTS and broker arrangements.

Edit: once an engine generates the trades, it goes into order management system (flowing through compliance). Order execution via Quantopian is only great for retail clients — not sure how many of these would be Q Enterprise customers.

My curiosity is the path for Enterprise clients to trade. Is this a use case being considered? Fawce's announcement above and these links don't actually explicitly talk about trading:

https://open.factset.com/catalog/products/quantopian-enterprise/en-us#
https://open.factset.com/catalog/partners/quantopian/en-us#

For example, these words are used:

The platform provides everything needed to research and develop effective quantitative investment strategies, supporting the workflow end-to-end, including research, alpha factor construction, portfolio optimization, strategy simulation, and evaluation.

What about trading? I guess if I were an Enterprise client, I'd want to go beyond evaluation, and have a path to generate a live feed, to plug into my trading platform. Maybe this is implied, and the Enterprise system would include real-time data feed support through the Portfolio Construction step in the workflow (leaving the Execution step to the Enterprise client)? Is this the concept?

If you're correct, then the purpose of the platform would solely be to harness algorithms for large institutions, while simultaneously denying those clever enough to develop great strategies the opportunity to profit from the same. There's also no guarantee given this business model that Q can be trusted not to sell your algorithm, or create something similar and market it to their clients. If by simply entering a contest, they have the capacity to view your algorithm, as it's put here, then your ideas are instantly exposed the minute they evaluate your algorithm. There are a lot of things here that do not add up. For instance:

  1. If you look at the leader board, the top earner has made $1000, but a professional python programmer can make that in his sleep. If you're actually smart enough to create something that an institution would find valuable, you're probably making a lot more than $50 and probably wouldn't waste your time.

  2. I entered one contest and made 6 million on 10 million invested over the course of the contest. I had a 10% draw down over the course of hundreds of trades. The winner of the contest averaged 3.5% yearly, but had a microscopic beta. That was about the only positive when the two were compared. I was also number two in Sharpe ratio. It may be that I'm missing something, but how in the world could you sell that to an institution when investing in the S&P500 would likely get you a 6% gain, inflation is 5%, and you could do just a well with a bond?

  3. Who are these Quanotpian winners who have received allocations? I can't find a single one.

All that said, I could care less! Let me trade my own algorithms, and I'll pay you for the use of the platform. Quantopian is nothing without its user base. Without us, you have nothing. Give us something in return for what you take and the ideas we expose to you. That is what the user base wants, not making hedge fund managers richer than they already are. Put a price of the feature for goodness sake. What do you have against more money? We're not all impoverished to the point we can't pay a monthly fee. If you want to draw the best developers to your platform and stop feeding them to QuantConnect, you have to give them something in return. A potential opportunity at $5000 isn't going to cut. It's actually insulting.

I'm wondering if this might be a path to revive some form of retail trading for the Q community? For example, I see that FactSet supports real-time data feeds:

https://www.factset.com/data/datafeeds/realtime

And then I see Quantopian and Interactive Brokers listed both as Channel Partners:

https://www.factset.com/about/channel_partners

So, I'm thinking that it is more-or-less a plumbing problem. What am I missing?

One potential opportunity for the Quantopian Community would be to contribute to the efforts of Quantopian Enterprise clients, and get paid for the work (e.g. consulting, coding, writing algos, etc.). Would it be possible to set up a kind of job board, so that Quantopian Enterprise users could post their needs?

Also, presumably, algos that are developed for the contest could be put up for licensing not only for use in the 1337 Street Fund, but by Quantopian Enterprise clients. Is this a use case you could support? It would be nice to be able to offer algos to the highest bidder, versus having them be exclusive to the 1337 Street Fund (as I understand, Quantopian Community members would still own the IP on their algos, so this would be a nice feature).

@Grant, that is a great idea. And it is where Quantopian should go, or at least explore the possibility, and get back to us on the subject. Quantopian has this huge talent pool (us) and is doing almost nothing with it except almost anonymize what we can do. Note that I am not for hire, but some might be.

It would be another way to incentivize strategy developers to participate in this community. As you must have noticed forum activity has decreased considerably over the last few months. An indication that maybe a lot of the talent has left Quantopian for other frontiers where they might find more opportunities.

I have to say, this is confusing. At one point, awhile back, I recall Fawce talking about aspiring to Quantopian becoming a $10B hedge fund, hired an industry veteran chief investment officer, quietly set up the 1337 Street Fund, hooked up with Steve Cohent/Point 72 (and reportedly still have a $250M commitment), dropped support for retail trading (bummer, but actually kinda made sense), and re-jiggered the contest to be in better alignment with the needs of the hedge fund (a no-brainer). So, I thought, "O.K. This sounds pretty focused. Quantopian is a crowd-sourced hedge fund." So, the move to Quantopian Enterprise and the SaaS model is kinda odd. A typical hedge fund, I would think, would not also provide their entire platform and resources to their competitors, but I guess if it is all gonna be out in the open on Quantopian Community, then why not also offer it as a service to institutional investors? The thing is, I figure that for Quantopian Enterprise to be an attractive product, clients need to be able to use it for doing exactly what Quantopian is doing, which is investing capital. So, in the end, wouldn't Quantopian Enterprise institutional investors be competing both for Quantopian Community members as employees, and in the marketplace for capital? I'm assuming that this has all been thought through and I'm just missing a certain perspective, but on the surface, it is very confusing.

For example, we have this statement:

We are committed to maintaining parity between the Quantopian Community and Quantopian Enterprise platform capabilities.

If I follow it to its logical conclusion, then a Quantopian Enterprise institutional client should be able to set up a crowd-sourced hedge fund in direct competition to the 1337 Street Fund, with all of the same resources, right? I suppose the costs might be higher, since they'd have to pay for the SaaS, but otherwise, there would be "parity." And if the 1337 Street Fund is wildly successful, then there will be plenty of competition signing up for Quantopian Enterprise. Why would Quantopian, in the end, want to support competition for its hedge fund (and indirectly, competition for Quantopian Community members, who receive allocations)?

@ Guy - Yes, the forum suggests a limited number of core users and a small trickle of new users. I get the sense that it has been de-prioritized within Quantopian, as well, based on their responsiveness and willingness to maintain conversations. And the contest only has ~125 entrants out of 190,000 registered users--not exactly a high participation rate. However, perhaps the algo quality is much better than for past contests, given the new rules.

To start, I want to see Q succeed as they have been an instrumental part of my personal development. I am spending the majority of my personal time elsewhere for a very simple reason: live trading. And I wish that wasn't the case because Q has a far superior platform to anything else I've seen.

When I saw the first sentence 'Sit down before reading' I got excited. Its amazing Q is adding this data/support, but I think a lot of us were hoping for live trading to be in the mix there following that statement. Awesome we get the tools but what's the point of building something without having a direct avenue to deploy it?

Just my opinion - without live trading, talent will continue to go elsewhere and/or use the platform here for personal research with little intent towards the contest or Q's main interests. I could be wrong. To me, democratizing trading means the full life cycle of a trader's journey. @ Aniken $kywalker made some good points above. If 99%+ of the community won't see an allocation then what's the incentive? At least with live trading, there are a few positives:
a) The hope to live off trading, even if it doesn't come to fruition. At Q it is a real achievable goal due to the abundance of resources.
b) IF live trading provides income, the ability to explore the contest and research full time/contribute to the community.
c) For spin-off funds, 3rd party IP sourcing, etc - live trading will provide real track records. i.e. some fund wants high returns and will sacrifice for higher volatility. Another way for Q to monetize off traders who wouldn't normally get an allocation from their internal fund...

Hope I'm not dwelling too much of sentiment from the fall and beating a dead horse. I'm excited to explore international equities with some of the algos I've developed - just wish there was seamless API connectivity like Q once offered to RobinHood and IB.

My theory of the Quantopian-FactSet partnership is that Quantopian is in it for the international data and FactSet is in it for the ability to offer Quantopian Enterprise. Notice that Quantopian Enterprise is not really a Quantopian initiative but a FactSet initiative; the links for Quantopian Enterprise go to FactSet's website. Quantopian Enterprise is what Quantopian can offer to FactSet in exchange for FactSet's international data. Within the FactSet ecosystem, the lack of live trading doesn't matter much; FactSet is an integrator that sells all kinds of data and tools to enterprise customers. One can infer from the FactSet catalog itself that those enterprise customers use a variety of tools and probably don't need a single tool (Quantopian) to be an end-to-end platform. Quantopian Enterprise is a way for FactSet to capitalize on Quantopian's popularity and expand the FactSet catalog. Meanwhile Quantopian maintains its existing focus on the hedge fund and community but now gets international data into the mix.

This theory obviates the confusion some people are having and allows each side of the partnership to be seen as continuing its existing business model trajectory.

Connecting the dots, it is reasonable that Quantopian Enterprise customers would be interested in real-money live trading. Above, in response to "Just out of curiosity, will Quantopian Enterprise include direct connection to the client's broker and automated execution of orders submitted by an algorithm?" Fawce said:

Quantopian Enterprise includes the research and development workflow only

So what is the path for a Quantopian Enterprise customer to use the Quantopian platform for trading? I've looked over various carefully crafted press announcements, and there is no mention of end-to-end usage of the QE platform, so the message is consistent. However, as we all know, Quantopian is not just a research platform for exploring historical data sets; it is capable of generating trading signals on a daily and minutely basis, and supports a hedge fund (kudos to y'all...nice work). So, wouldn't QE customers want to leverage the full capability of the platform? Perhaps the idea is to work with QE clients on a case-by-case basis, and address the need for live trading as it arises? Or from the get-go, just tell them it won't be supported, and have them check out zipline-live?

Then, I'm thinking...if QE (eventually) supports live trading, couldn't an enterprising Enterprise customer figure out how to offer live trading to the Quantopian Community? What would be the impediments (aside from a limited number of users...which perhaps is the most fundamental problem)?

On https://open.factset.com/helpdesk/faq/en-us#faq-data-providers, it sounds like FactSet is open to the idea of adding new data providers:

If you are interested in becoming a provider, please email [email protected].

I'm wondering if this also might be a Quantopian Community use case? Say one created an algo that provides a data feed of interest to institutional investors. Would it be something that could be fed into FactSet? Would the Quantopian/FactSet partnership support this?

Didn't know FactSet is also a signal provider.

I would have thought they'd be after data (e.g. banana production and quality).

Hi Fawce and Josh -

Any thoughts on the ideas above? I think there might be some interesting discussion points for the community.

Thanks,

Grant

The irony is rich here... John makes this momentous announcement, but the majority of discussion appears to be of disappointment in the continued lack of live trading support.

From my soap box here, I don't see live trading support ever coming back to Q (at least not in the near future). Q's core business isn't providing you with a product or service. It's creating an environment that causes you, the community, to compete for their business. You provide them with your intellectual property, and they compensate you with a small cut of the profits it generates. If you use their platform for any reason other than to enrich the quality of competition for their business, they don't care about you or want you. It's not because their some evil enterprise... it's just that business is business.

Remember folks, they are a hedge fund. Their idea was brilliant. They created an environment for quants that allowed them to poach incredible talent for a tiny fraction of what their services would go for at any other firm.

How much do you think one successful algorithm is worth to Q? Their bottom line is just fine without trying to be a tech company.

If anything, I could see them creating a separate platform that falls under Q's holdings for paid live trading services. That probably wouldn't be for some time though. There's more money elsewhere right now. As anyone would, they're going to harvest the low bearing fruit first.

Also, if anyone is interested in collaborating to create a live algorithmic trading platform similar to old school Q, shoot me an email at [email protected]! I would love to work with you! With the business model I've formulated, I'm confident the platform could yield fairly substantial profits and still be either free or very reasonably priced.

Also, sorry for being a dirty double poster. I hope you, dear reader, can find it in your heart to forgive me ;).

Great news ! Is there any chance Q provide Hong Kong stock market and fundamental data? There are so many platform like Quantopian in China (not good as Q), but just one company provide HK stock market. But even that, they provide HK data after 2015.

Hi Stanley,

We don't yet have HK data integrated on the platform, but we are planning to add it as we add global equity markets. You can read a bit more about our plans here.

Disclaimer

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by Quantopian. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. No information contained herein should be regarded as a suggestion to engage in or refrain from any investment-related course of action as none of Quantopian nor any of its affiliates is undertaking to provide investment advice, act as an adviser to any plan or entity subject to the Employee Retirement Income Security Act of 1974, as amended, individual retirement account or individual retirement annuity, or give advice in a fiduciary capacity with respect to the materials presented herein. If you are an individual retirement or other investor, contact your financial advisor or other fiduciary unrelated to Quantopian about whether any given investment idea, strategy, product or service described herein may be appropriate for your circumstances. All investments involve risk, including loss of principal. Quantopian makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances.

Thanks, good to know.

We believe you can be more successful with more markets and more data

What do you mean "more succesful"? More successful in creating you algos as live trading is no longer supported?

This is now a hobby site for hobby traders.

Congratulations team!

I hope this could help bringing quality fundamental database, so value investors could dig without finding missings, and developing strategies.

Morningstar is suppose to be as famous as FactSet, so im afraid of having similar fundamental data issues. Anyway, we must give it a chance.

Really looking foward to use this Dataset.

Thanks

Hey Santiago,

FactSet is widely used in the asset management world, especially for quantitative management, while morninstar is not (at least not at the same level), So that would surprise me that there were similar issues in the data quality provided by FactSet

@Mathieu, @Santiago,

Regardless of whether FactSet or Morningstar fundamental data is used, the source (for the most part) has the same origin, corporate filings (10Q & 10K) with the SEC or for securities outside the US, their respective regulatory bodies. The more important issue is how well these data providers structure, process, clean and integrity check these raw data into something that is meaningful for quants to digest in their search for the elusive alphas.

My concern for risk conscious Quantopian is that they are just passing through the data from these third party providers to us without ( or with very minimal) data integrity and filtering of their own and this and by itself poses a risk to their allocation process.

Let me give a concrete example. Hypothetically, if in one of my algos, I used peg_ratio, initially unaware that Morningstar only started providing this data in 2014. I ran this multi-factor algo with peg_ratio from 2016 to present and get good results (risk adjusted returns) . I also ran my holdout data from 2008 to 2015, and again get good results. If I entered it in the contest and it performs well enough for Q to consider it for allocation, how then will Q, who presumably does not look into your code, know about this missing data anomaly of peg_ratio when Q does not have complete management control of these input data. Isn't this a risk by itself, unless I'm missing something?

The solution I believe is for Q to do due diligence and management on data they receive from their third party providers regardless of their reputation in the industry, most specially in the context of their contest format which eventually leads to fund allocation. They have done it with filtering QTradableStocksUS and contest constraints and thresholds. All Q needs is to devote some resources in doing filtering, cleansing, structuring of these fundamental data into a "QFundamentalUS" inputs for their own sanity checks and thereby mitigating their risks of exposure to bad, incomplete, unrealible fundamental data.

Any thoughts?

@ James -

One of the many mysteries of the Q. They've never taken data integrity seriously. If they did, there would be a public-facing data base of any and all reported and verified bugs/glitches/anomalies, etc. And better yet, the data base would be accessible within algo code (e.g. of there's something wonky about a given data set, then it could be avoided until it has been flagged as clean).

Maybe Q will turn over a new leaf in this area, with the advent of Factset and the SaaS business model for institutional researchers?

If I entered it in the contest and it performs well enough for Q to consider it for allocation, how then will Q, who presumably does not look into your code, know about this missing data anomaly of peg_ratio when Q does not have complete management control of these input data. Isn't this a risk by itself, unless I'm missing something?

My hunch is that there is wiggle room in the terms of use for Q to look at both inputs and outputs ("exhaust") of an algo. Additionally, if one signs a contract, then I'd have to figure that there may effectively be an addendum to the terms of use, that while not going to the point of accessing the code itself, anything necessary for risk management would be allowed, including monitoring inputs (seems reasonable, when up to $50M could be on the line). Or perhaps the terms of use don't apply to algos licensed for the fund, and what's in the contract is all that matters?

Anyway, the idea of not looking at code, and only "exhaust" seems like an untenable plan anyway, long-term. At one point, Fawce talked about having $10B under management. Doing this with black-box algos...not so realistic, I would think.

@Grant,

Hunch? I'd like to hear it directly from Q management because as I alluded to, this poses a risk and I would like to know how exactly they handle this. This seriously could affect their bottom line, not that we will ever actually know their actual fund performance, it's just between Q and Point72. But I sure do hope that Q addresses this issue. The solution is already there, the process of filtering QTradableStocksUS is a good example, they just need to devote resources to it and with the entry of Factset and their enterprise roll out, they should do it for their own sanity and confidence.

@ Fawce -

One curiosity is what backtesting interface will be used by your Enterprise customers? On https://www.quantopian.com/factset, it is mentioned:

Log in and get started immediately with Quantopian's Python research environment and FactSet's data

Presumably, by "research environment" you mean the whole of Quantopian, as it is available to the Community, and not just notebooks. My confusion is that the backtester is narrowly geared toward the present Q contest/fund "constraints" whereas I would expect that your Enterprise customers would want a more general purpose interface.

One way to address this would be to provide access to the backtesting engine from within a notebook. Have you considered this? I think it would be a good option for the Community, as well, particularly if backtests could be launched programmatically in parallel, with select results returned (not the full backtest results, which require a ginormous amount of memory).

On a separate note, I would think Enterprise customers would need is a full workflow path to trade real money (at least on a pilot level). You could take the approach of "We'll cross that bridge when we come to it." The issue I see is that one would not want to go through a development cycle on Quantopian, and then need to port the algo over to another platform. Eventually, are you planning to offer a full workflow to your Enterprise customers?

One suggestion to drive out data integrity issues would be to include them in the Quantopian bug bounty program (which has paid out $51,735 total so far...holy cow!). It appears to be geared toward security issues, but it could be extended to include any sort of not-workiness on the platform.

Hello everyone,

As far John details about the duality of FactSet data, it seems Quantopian Enterprise is already working, offering a platform integrated with FactSet. However, is it possible to access to the FactSet in Quantopian Community?
Maybe im wrong, but i havent seen any algorithm using FactSet and either any tutorial of how to use it in the current platform.

Does anyone know about this integration process?

Thanks,

Hi Santiago,

There hasn't been any further announcements yet as far as I know. I believe they are currently working on integrating the FactSet datasets into the Q platform, in parallel with rolling out other global equity markets as well.

Perhaps an update from the Q team (ideally with some rough timelines) might be due?

We are currently working on integrating FactSet's fundamental data. In parallel, we are working on adding data from a new market to Pipeline. We're hoping to have one or both of these available to use on the platform in the coming weeks. We'll make the announcement in the forums as soon as the new data is available!

Awesome! So looking forward to both!

Will the the FactSet Fundamentals and new market be available for use in the existing contest, or will there be a separate contest for the new market? E.g. a Q US Open & a Q [new market] Open? Will the new market be based in the local currency or how will currency hedging work? (I can’t think of a USD denominated equity market outside of the US)

@Joakim, these are great questions, thanks for asking.

FactSet Fundamentals for US equities will be usable in the contest. International market data will be available in Pipeline only (i.e. you can build and test factors in research, but you won't be able to backtest with it yet). As such, international market data won't be usable in the contest right away. I hope to have more info on the scope of the international market data once it's been added to the platform.

Regarding currency, the data will be based in the listing currency of each asset (this is usually the same as the local currency). At some point, we will probably look to add an option to convert price-based values to a base currency, but it's not yet at the top of our list.

P.S. If you have any follow-up questions pertaining to the pending global equity expansion, do you mind asking them on this thread? I think it will be easier for other community members to track if we consolidate the questions in one place (I'm not just trying to promote my ghost of a thread, I swear).