Crowdsourced alpha and futures contracts on data. Five fintech startup ideas inspired by quantamental trends in the investment management.

In 2020 retail investors set new records of stock market participation, artificial intelligence continues transforming investment research, investment firms rely on alternative data in an environment of extreme uncertainty, and even the most secretive and conventional financial institutions have to face the new reality of managing remote workforce.

All these major changes are also monetizable opportunities for fintech entrepreneurs.

1. Best investment ideas platform-alpha capture meets social trading.

Automated alpha capture systems enable professional research analysts who work for investment banks and broker dealers to submit trading ideas in a standardized electronic format. Buy-side firms pay for those ideas based on the analysts’ ratings. Submissions include investment rationale, timeframe and conviction level, and the platform enables investors to quantify the performance of trade ideas.

On the other side of the spectrum, there is social trading — social networks for individual retail investors where they share their trades ideas, discuss and follow other people’s strategies. And with the rapid growth of retail participation in the stock market, the popularity of these websites and apps also increases.

The idea is to create a platform with a top quality UX and elements of gamification for individual traders that could quantify ideas’ value and facilitate fair performance based compensation for the contributors’.

Of course, you may think, there are tons of social trading apps and forums. I agree, and it only proves the concept. Some of the most popular platforms for investment ideas exchange look like websites from 1990s and there is a room for improvement. Besides, Robinhood didn’t invent retail trading either, yet it keeps outrunning established rivals.

Individual investors’ trading ideas and more importantly users’ reactions to them would make a great dataset that professional investors would be willing to pay for, specially now when Robintrack, the most popular source of data on retail positions is about to shut down.

2. Crowdsourced expert data labeling.

Data labeling is one of the most important factors in building effective deep learning models. Accurately labeled data can provide ground truth for testing and iterating AI models.

You know how when fill out a form on the internet, you’re occasionally asked to identify a bus in five different pictures to prove you’re a human. This is a so called “Captcha” (Completely Automated Public Turing test to tell Computers and Humans Apart)? Captcha is an elegant tool for training AI algorithms, for example in image recognition. And there are platforms that offer captcha entering jobs, they pay users a very small amount of money to get tons of people involved in manual labeling.

When big data is used for investment analysis, linking the data points to companies and exchange tickers is curtail. Of course, hiring people to do this manually isn’t practical and when this is done automatically, oftentimes it creates too many errors. Another example of the importance of data labeling is in tagging market moving news as positive/neutral/negative. It is currently mostly done by NLP (natural language processing) algorithms based on dictionaries. But humans can rank news and discussions much more accurately to identify hints, expressions, emotions, sarcasm, and other less than obvious patterns.

The idea is to create a platform that awards users for ranking market moving news, blog posts, twits and linking news/posts to companies’ identifiers. Financial companies would love this sort of datasets, specially if the user base includes individual investors who understand the market and follow news sources. Once the platform gets to a certain point in terms of its user base, it will become a powerful tool to apply to endless number of datasets, not only in finance, but also in healthcare, ads, consumer goods and services and more.

3. Process oriented alternative data collector.

There is a common belief that alternative data is supposed to be very exotic and very exclusive, but in fact free data from publicly available sources can be of a great interest and a great value for financial firms.

A good example of that is Robintrack, a website created by a student as a college side project, that is now extremely popular among multi-billion dollar hedge funds. Another example is publicly available regulatory filings in the EDGAR database, that are one of the most popular and frequently used sources of data for hedge funds.

There are many data types that can be collected online for free. But it takes time and effort to process the data into a usable format and financial companies will be happy to pay for a ready to use data product with a few years of history. All it takes is to consistently collect and store data with timestamps and adjustments, and carefully document the methodology. Once the initial setup is done, it is a fairly passive process and you can move on to the next data source.

Such a business model of a nimble data collector is more viable than vendors that depend on just one data source or aggregators that merely connect buyers and vendors. Over the years you may end up having a huge competitive advantage with high quality data history. And if something goes wrong with the data source, you’re diversified in terms of data sources and client base. The key to competitive edge is to understand clients needs and data engineering process.

Selling alt data may seem like too much of a niche business to become a scalable fintech start up. But in a few years when alternative data gets democratized and becomes available in a more affordable and usable format to passive investment sponsors and even individual investors, such a data collector will easily become another billion dollar valuation company.

4. Freelance platform for quant analysts

Hedge Funds like Quantopian and WorldQuant have been hiring remote quant researchers for years. And in the current environment for obvious reasons such a crowdsourcing model becomes relevant to virtually every asset management firm.

And there is an opportunity in creating an independent centralized freelance platform (sort of like Upwork) for quant researchers.

  • Employees don’t need to be physically present at the office these days. Thus, financial firms no longer need to limit their hiring strategy by a certain location;
  • It became close to impossible to arrange working visas for foreign employees, so the only way to access global talent pool is remote work;
  • Many funds have made layoffs and may not be ready to hire full time, the flexibility of a freelance arrangement makes this model more cost efficient;
  • Using crowdsourcing hedge funds get the benefit of a wider pool of potentially uncorrelated strategies;
  • Freelance projects allow quant researchers to showcase their strategies, build a track record and find full-time opportunities in hedge funds and sell side firms.

Such a quant freelance platform would need to facilitate contractual relationships, non-disclosure agreements and the compensation structures.

The important factor that such a platform would need to figure out is access to data. One way to do it is for the platform to purchase enterprise licenses from alt data vendors so that freelance quants can use it within the platform to develop and test strategies. Then platform clients (hedge funds) would need to have their own subscription to independently verify models and use them going forward. This way the platform could become a distribution channel for data providers.

Both freelance platforms and crowdsourced alphas are proven business models. And with the trend towards quantamental research in the investment industry combined with the overall global trend of remote work and gig economy, freelance quant researchers platform is another potentially scalable fintech startup idea.

5. Big data exchange

They often say data is becoming a commodity. If so, it would be natural to start trading data as a commodity. And I’m not talking about multiple existing market places and aggregators that act as a shopping window for existing data products. I’m talking about an actual exchange which requires commodity to meet certain quality standards, that regulates contractual responsibilities of buyers and sellers, and facilitates transparent price discovery based on supply and demand.

One potential feature of the exchange, could be deliverable futures contracts to purchase data at a specified price in the future (e.g. when enough history is available). This way the buyer can lock in the price, and the seller gets paid upfront. If between futures transaction and expiration date the buyer no longer wants the data, they can sell the contract on the secondary market.

Another potentially valuable feature is escrow accounts. If the buyer enters into a long term contract, the exchange can hold escrow accounts to only release payments if the service continues and the quality of data continues to remain the same. Otherwise, the exchange would facilitate cancellation for the breach of contract.

Such a big data exchange is by no means limited by the investment management use case. Investment banks, commercial banks, consumer goods corporations and media companies could be potential paying clients and equity investors for such a venue.

Capital markets fintech startups may seem less scalable and thus less interesting to VCs. But in fact, if you build a really useful tool and figure out a viable business model, not only it may become sufficiently scalable, you’ll get access to an additional group of investors — large hedge funds and investment banks who have separate divisions to make private investments in promising fintech startups.

Managing Director at