At IntoTheBlock, we have been working on predictive models for different crypto assets. The work ranges from sophisticated quant strategies to more basic directional forecasts. Based on that work, we have learned quite a few lessons about effective methods, challenges, and very peculiar aspects that should be taken into consideration when attempting to build predictive models for crypto assets.
In this article, I would like to explore some of those ideas that, hopefully, shed some light about the magnitude of the challenge of predicting the price of crypto.
“All Models are Wrong but Some Are Useful”
British statistician George Box once said that when it comes to statistical models “all models are wrong, but some are useful”. That phrase has been adopted as a mantra in quantitative finance as an indication that the ever-changing nature of financial markets will cause problems to the most sophisticated predictive models. In the case of crypto assets, this idea couldn’t be more correct.
The dynamic nature of crypto assets, the regular volatility ( yes, it will come back 😉 ) and the short trading history makes crypto particularly challenging for predictive models. Additionally, different crypto assets can be based on fundamentally different protocols and could behave differently during certain market conditions.
In the context of crypto assets, quants or data scientists working in predictive models should realize that even effective predictive models would have a limited time-span and would be vulnerable to changes in market conditions. From that perspective, it is more important to produce a variety of predictive models for different thesis than trying to nail the perfect predictive strategy.
The Case for Deep Learning in Crypto Asset Predictions
There are many methods that can be used to model predictive behaviors in crypto assets. Taking some liberties, most of those techniques can be grouped in one of the following categories:
- Time-Series Forecasting: Traditional statistical methods that focus on predicting a value in a time series based on existing attributes.
- Machine Learning: Simpler machine learning techniques such as linear regression or decision trees which are very common in quant strategies.
- Deep Learning: Predictive models based on the new school of deep neural networks.
In our experience applying time series forecasting methods to crypto-assets showed that, although they are relatively easy to use, they are not very resilient to the constant fluctuations of the crypto space.
Traditional machine learning methods such as linear regression or decision trees have a strong presence in traditional quant systems and, therefore, one might seem inclined to extrapolate those lessons to the crypto space. However, we found that those methods have a strong challenge generalizing and haven’t proven to be very resilient to market conditions.
Deep neural networks is the latest trend in the artificial intelligence(AI) space and the one that has been developing the fastest. In the last few years, the research body in deep learning has increased drastically. The promise of deep neural networks is that they can uncover complex non-linear relationships between arbitrary variables. In our experience, deep learning models in crypto assets have shown strong resiliency to market conditions and the ability to generalize knowledge. The biggest drawbacks are that these types of models are computationally expensive to produce and very hard to understand and interpret how they produce decisions.
In quantitative finance, the application of deep learning methods is still in very nascent stages compared to other methods. However, the promises are tremendous. In the case of crypto assets, deep learning models exhibit some tangible benefits:
- Complex, Non-Linear Relationships: Deep neural networks can model very complex non-linear relationships between different predictors.
- Resilient: Through constant training, deep learning models have proven to be resilient to the constant fluctuations in the crypto market.
- Large Research Body: Deep learning research in quantitative finance is growing faster than any other discipline providing lots of research ideas that can be adapted to the crypto space.
At the same time, there are a couple of potential drawbacks of applying deep learning models to crypto-asset price forecasts:
- Interpretability: Deep neural networks are complex and, therefore, really hard to interpret.
- Expensive to Build: Deep learning models are computationally expensive to build and maintain.
From the different statistical and machine learning schools in the current markets, deep learning seems particularly well equipped to handle the challenges of building predictive models for crypto assets. However, assuming that most deep learning ideas from traditional capital markets will apply in the crypto space would be a mistake. We certainly experimented with a few of the most accepted deep learning methods in traditional quant finance and encountered a few surprises.
10 Things We Learned While Building Predictive Models for Crypto Assets
Building a robust prediction pipeline is a very difficult task. Many of the models that work great in a lab environment will result hard to operationalize. Here are some lessons that might be relevant when considering applying deep learning models to crypto asset predictions:
- Training Size Matters: When it comes to crypto-assets, the larger the dataset to train models the better.
- Data Quality is a Huge Problem for Crypto Assets: It’s very difficult to find reliable data sources in the crypto space.
- Accurate Predictions do not Mean Actionable Predictions: An indication of directional movement in price is not a trading strategy.
- Operationalizing Real-Time Predictions are Hard: Running deep learning models in real-time require quite a bit of infrastructure.
- Blockchain-Based Predictive Models are Very Fragile: Blockchain predictive models are very vulnerable to exchange manipulations, forks and other runtime changes that can affect their performance.
- Deciding When and How to Retrain Models is Challenging: Retraining predictive models after they are in productions can change their performance in unexpected ways.
- Rapid Experimentation is Key: Building predictive models for crypto assets is a business of failure. Trying different ideas rapidly and iterating is essential for success.
- Data Sources in Crypto are Very Unreliable: Exchanges failures, missing data, wash trading records are some of the elements that affect the reliability of data sources used in crypto predictive models.
- Better Infrastructure Beats Better Models: A robust infrastructure with mediocre models will beat a poor infrastructure with great models in the long term.
- Most Quant Research Fails When Applied to Crypto: Most techniques in research papers were not designed for the dynamics of crypto markets.
Quant strategies are likely to be the dominant form of investing and trading in crypto-assets and, as a consequence, predictive models are likely to play an important role in the evolution of those strategies. The ideas in the field of deep learning applied to crypto are still in a very early stage and there is a huge gap between research and practical applications. Some of the ideas outlined in this article attempt to provide a practical perspective on the challenge of building predictive models for crypto assets. The challenges are many, but the journey is certainly fun.
About the authors
Jesus Rodriguez is the CEO-CTO of IntoTheBlock, a market intelligence platform for crypto assets. He is a computer scientist, a speaker, and author on topics related to crypto and artificial intelligence.
Lucas Outumuro is a Sr. Researcher at IntoTheBlock, a market intelligence platform for crypto assets. His areas of focus include crypto derivatives, DeFi and web 3.0 in general.
Like what you see? Subscribe for daily updates.
A bankers guide to AI Part 3. Does the AI have more than one purpose? What is the roadmap?
This is the third in a 5 part series (published weekly) written by guest author Amber Sutherland a banker who understands technology who currently works for Silent Eight an AI-based name, entity and transaction adjudication solution provider to financial institutions. Click here for Index and Part 1.
Many financial institutions have the dueling mandates to be both innovative and transform digitally, but also to rationalize vendors. So, when considering artificial intelligence solutions, which are often niche, it’s worthwhile finding out:
- How the vendors decide to build out features;
- Whether they are willing to customize their offering for you;
- How reliably they’ve delivered on features in the past; and
- Whether what’s on their roadmap adds value for you.
This way you can ensure that the decision you’re making is one that is future-proofed and set up for longevity.
Stay tuned next week for Part 4. Is it better than what you have now?
Daily Fintech’s original insight is made available to you for US$143 a year (which equates to $2.75 per week). $2.75 buys you a coffee (maybe), or the cost of a week’s subscription to the global Fintech blog – caffeine for the mind that could be worth $ millions.
Stablecoin News for the week ending Tuesday 11th August
Here is our pick of the 3 most important Stablecoin news stories during the week.
Russia’s bill to regulate cryptocurrencies has been signed into law by President Vladimir Putin. The new law gives legal status to cryptocurrency but prohibits its use as a means of payment.
The law provides a definition to digital currency, stating that it “is recognized as an aggregate of electronic data capable of being accepted as the payment means, not being the monetary unit of the Russian Federation or a foreign state, and as investments,” Russian news agency TASS described. “The digital currency cannot be used at the same time to pay for any goods and services.”
Meanwhile, the law sets forth that digital financial assets “are digital rights comprising money claims, ability to exercise rights under negotiable securities, rights to participate in equity of a non-public stock company and right to claim transfer of negotiable securities set in a resolution on the DFA issue,” TASS noted. These assets can be sold, purchased, exchanged, and pledged. However, they cannot be used as a means of payment.
Russian banks and exchanges can become exchange operators of digital financial assets provided that they register with the central bank, the Bank of Russia.
Almost the very next day up pops SberBank, the largest and state owned retail Bank in Russia, with advanced plans to use Hyperledger for it’s very own stablecoin.
Meanwhile, China’s big four state-owned commercial banks have started large-scale internal testing of what would be the world’s first sovereign digital currency, as the launch of the digital yuan appeared to move a step closer, the 21st Century Business Herald reported on Thursday.
The Bank of China, the China Construction Bank, the Industrial and Commercial Bank of China and the Agricultural Bank of China are working on the digital yuan with the central bank in major cities, including Shenzhen, according to the Guangzhou-based newspaper.
Users taking part in the trial can use the app to top up their accounts, withdraw money, make payments and transfer money after registering with their mobile phone number. The banks are also testing a scenario where a user can make a transfer to another account without an internet connection, the newspaper added.
Another interesting angle on the Chinese CBDC from the FT, they speculate that in the past, the People’s Bank of China (PBOC or Central Bank) gave the local Tech giants an easy ride in the payments space and is now looking to balance things going forward.
So what we are seeing is that these two very centralised State actors are looking at CBDC’s as an opportunity for the Central Banks to give the State Banks a major leg up. Will the west follow or go a very different way?
Alan Scott is an expert in the FX market and has been working in the domain of stablecoins for many years.
We have a self imposed constraint of 3 news stories per week because we serve busy senior Fintech leaders who just want succinct and important information.
New readers can read 3 free articles. To become a member with full access to all that Daily Fintech offers, the cost is just US$143 a year (= $0.39 per day or $2.75 per week). For less than one cup of coffee you get a week full of caffeine for the mind.
Streamlining manual processes amid COVID-19
The efficiency of any operations department within financial services depends on workflow, especially when working remotely during the coronavirus crisis, and ultimately, efficiency is a key indicator of bottom-line profits.
To streamline work processes and improve workflow, a business must have an overall assessment of operations. Financial institutions must evaluate office paper use, including the usage, processing and archiving of forms. The financial services industry must look for areas of improvement that move towards digitising manual processes and eliminating physical paperwork from order forms, onboarding documents and agreements.
DeeDee Kato, vice president of corporate marketing at Foxit Software tells FinTech Futures about what banks can do to ease the process of digitising paperwork during the coronavirus crisis and the challenges faced with remote working. From scanning documents, using optical character recognition (OCR) to make it text scannable and searchable, to the issues people face when relying on paper located in the office.
“The number one thing that’s the most important factor for our banking customers is customer service, as there are many things to think about when going digital, especially during the pandemic since there’s a whole uptick in online activities,” says Kato.
Statista’s July 2020 online activity report shows that “almost 4.57 billion people were active internet users as of July 2020, encompassing 59% of the global population”. With such high internet usage comes the need for banks to ensure they can support external parties and internal teams whilst working remotely. “That’s where streamlining workflows with robotic process automation (RPA) and machine learning can improve productivity,” adds Kato.
Gloria Sánchez, group head of legal for technology and legal transformation at Banco Santander, highlights the need to digitise cumbersome paperwork in a legal setting amid COVID-19. “The legal department tends to be quite traditional,” says Sánchez, “the pandemic sped up the need to digitise, although we were already in the process of being fully digital before COVID-19, there is still a lot of work to do within the implementation processes.”
Sánchez tells FinTech Futures about the importance of improving processes when providing information to regulators. “You may have the information in many documents and many repositories when applying data structure techniques, so you need OCR.” This creates a bottleneck as compliance teams must search across departments and verticals in the group to obtain the relevant information for the regulator, an often time-consuming process.
Kato mentions that Foxit’s OCR capability solves the problem of static images to making them text scannable, especially when handling multi-page documents converted to text files. A customer can use keyword searches to find critical information and copy and paste values into the regulator’s system.
“Regulators often receive these document requirements as paper documents and they mention how they need to make comments and remarks on them and send them to another party, who also had to make a paper copy, which is quite cumbersome,” notes Kato, “but now they scan it and apply OCR so it can contain legitimate texts which can now be edited and annotated.”
Some of the areas for both Kato’s clients and Sánchez’ teams that have moved towards digitising their processes are know your customer (KYC) management, converting mortgages and loans documentation, forms processing, compliance and security, signatures, archiving and retention and mobile apps/online banking.
“Our goal is to enable our customers to go fully paperless in each of these areas to significantly reduce operational costs associated with paper and manual processes,” says Kato.
During the KYC process, banks receive a multitude of files and documents from different clients. These entail identification documents, proof of income, proof of residence and more. The issue arises when the documents arrive in various formats such as scanned images, skewed scans, photo images, they all need to be standardised to be easily searchable and retrievable.
“Whilst the goal for banks and credit unions is customer convenience, it’s not usually the case the other way around. Customers can send it through a scan or take a picture of it and send it that way, scanned images, large text files, which are less than ideal,” notes Kato. An influx of these more static files only increases the KYC or regulatory review process as the documents require remediation before use.
“By converting the documents to PDFs or make use of digital portfolios with PhantomPDF, you can reprocess the documents and keep the original file,” adds Kato. This makes collecting and retrieving client information a lot easier for employees working on KYC procedures. For badly scanned images or file types that are way too large, PhantomPDF can automatically re-adjust the image, compress, and convert it to PDF while maintaining the original integrity of the document.
“Firms need to re-think and overcome the challenge of heavy paperwork and manual processes, so they need to meet the customer where they live,” says Lil Roberts, CEO at Xendoo. “They live in text messaging, mobile, live chats on website, emails and phone calls. But what the problem has been in the US is that practitioners don’t value the customer service side of the business and taking care of small business owners in a timely manner; customers need and want swiftness.”
Banks are incredibly reliant on physical documentation when reviewing complex cases like mortgages and home refinancing. Prospective clients submit a lot of documentation such as mortgage payments, tax returns, statement of assets, identification and more.
“We’re talking about customer documentation that needs to be deeply reviewed as part of the KYC procedure and in the situation we’re in with fluctuating rates, income loss caused by the coronavirus crisis combined with customers preferring to minimise branch visits, we foresee challenges in managing an influx of documentation which never comes in neatly organised or in standardised format and the process can be dragged out,” says Kato. “We’re the solution to this problem.”
Foxit’s PDF technology enables banks to not only transition from paper to paperless but also gives them the ability to work with digital documents like they would with paper.
Visit Foxit Software for more information on PDFs.
SEMI Partners with GLOBALFOUNDRIES to Offer Apprenticeship Program Aimed at Building the Electronics Talent Pipeline
Materials science researchers develop first electrically injected laser: The diode laser uses semiconducting material germanium tin and could improve micro-processing speed and efficiency at much lower costs
Advance in programmable synthetic materials: Reading sequence of metal atoms in MOFs allows encoding of multiple chemical functions
Invest 3% in Bitcoin to Avoid COVID-19 Lockdown Devaluation — BitGo CEO
Cointelegraph Launches Newsletter for Professional Investors
Bitcoin Cash short-term Price Analysis: 12 August
Token Launches From Ethereum to Telegram: Where Do We Go From Here?
Enterprise VR Hardware Specialist Varjo Raises $54 Million in Latest Funding Round
Grayscale Bitcoin Trust Saw Surge in Investor Interest After March
VeChain & Oxford Announce New Framework to Assess Consensus Protocols
Championing Blockchain Education in Africa: Women Leading the Bitcoin Cause
Evening Reading – August 11, 2020
Chainlink: Traders under zero loss, but why?
The Babylon Project: A Blockchain Focused Hackathon with a Commitment to Diversity & Inclusion
Varjo Raises $54M Financing to Support Its Retina-Quality VR/AR Headsets for Enterprise
Ethereum, Zcash, Dogecoin Price Analysis: 12 August
Peer-to-Peer Exchange CryptoLocally Now Offers Instant Credit Card Payment
Cardano (ADA) Holds On to Crucial Support By a Thread
Bitcoin Creates Double-Top After Failing Close Above $12,000
DeFi Farmers Rush to Yam and Serum for Explosive Yields
Copper Foil Market Size Worth $10.3 Billion By 2027 | CAGR: 9.7%: Grand View Research, Inc.
Corundum Market Size Worth $3.5 Billion By 2027 | CAGR: 4.0%: Grand View Research, Inc.
Mozilla is Shuttering its XR Team Amidst Major Layoff, But ‘Hubs’ Will Continue
New Energy Challenger, Rebel Energy, Places Blue Prism Digital Workers at the Heart of its Launch Plans
Teknosa grows by 580 percent in e-commerce and pulls its operating profit into positive territory in Q2, despite the pandemic
Novo Ventures Portfolio Company F2G Closes US$60.8 Million Financing
F2G Closes US$60.8 Million Financing to fund late stage development of novel mechanism antifungal agent
LocalCryptos Integrates Inbuilt Crypto-To-Crypto Exchanges, Powered by ChangeNOW
Putin’s plan for Russia’s coronavirus vaccine is at ‘high risk of backfiring,’ expert says
UK enters recession after GDP plunged by a record 20.4% in the second quarter
Another Steam Game Festival Is Coming In October
Top 25 Nationally Ranked Carr, Riggs & Ingram (CRI) Welcomes Cookeville-Based Firm, Duncan, Wheeler & Wilkerson, P.C.
Avast plc Half Year Results For The Six-Months Ended 30 June 2020
Russian hackers steal Prince Harry and Meghan Markle photos via Cyber Attack
Oddworld: New ‘N Tasty Coming To Switch In October
Linkin Park’s Mike Shinoda Is Writing A Song For Gamescom 2020
Texas School District experiences DDoS Cyber Attack
‘EVE: Echoes’ from CCP Games and Netease Is Now Available Early on the App Store, Servers Go Live Tomorrow
Hans Zimmer Created An Extended Netflix “Ta Dum” Sound For Theatres
Everything you need to know about the Exxus Snap VV
Gaming6 days ago
Server status – Is Fall Guys down?
AI1 week ago
How to Purchase a Future Tech Company
Esports1 week ago
The best loadouts for the ISO in Call of Duty: Warzone and Modern Warfare
AI6 days ago
AI Machine Learning Efforts Encounter A Carbon Footprint Blemish
AI1 week ago
Playstation Assist- Sony’s brand new smart assistant
AR/VR1 week ago
Extended Reality- Immersive Experience
Automotive1 week ago
What did you do during lockdown? Right, you were playing ‘Grand Theft Auto’
Esports1 week ago
The best Standard Hearthstone decks to try for Scholomance Academy