Client Story

Advancing anti-money-laundering detection

KPMG helped a US-based cryptoasset exchange build a scalable and reliable AML transaction monitoring platform to automatically identify suspicious activity among billions of transactions.

John Caruso

John Caruso

Principal, Forensic, KPMG US

+1 212-954-6831

Thomas P. Keegan

Thomas P. Keegan

Principal, Advisory, Forensic, KPMG US

+1 212-954-7880

Sal Ternullo

Sal Ternullo

Director, Advisory, U.S. Cryptoasset Services Co-Lead, KPMG US

Client
A U.S.-based cryptoasset exchange
Sector
Financial services
Project
Anti-money-laundering transaction monitoring platform
  • Client challenge
  • Benefits to client
  • Approach
  • KPMG insights

Client challenge

Regulations surrounding digital currencies or cryptoassets hold similar expectations to traditional currencies. For example, the requirement to have a system in place to monitor all transactions customers perform and identify anti-money-laundering (AML) activities is similar for both traditional financial service institutions and cryptoasset companies. However, until recently, there has not been a commercially available system to do so on a large scale. Even worse, the systems in use were generating a considerable number of false positives—flagging transactions as money laundering activities that actually were not. This triggered time-consuming, manual verification processes to clear the alerts.

When state regulators inspected the AML capabilities at a leading U.S.-based cryptoasset exchange—doing billions of transactions every month for millions of global customers—the company knew it needed a better solution, and fast.

Benefits to client

KPMG was engaged to help our client leverage our advanced analytics capabilities and cryptocurrency experience to advise our client as they design and implement an enterprise-wide AML transaction monitoring platform. Utilizing this platform, the company is now able, for the first time, to:

  • process billions of traditional and cryptocurrency transactions, and automatically monitor a variety of AML scenarios at the same time to identify potentially suspicious activity anywhere across the enterprise
  • reduce the number of false positives
  • considerably reduce the level of effort to manually verify the accuracy of AML alerts 
  • work with one primary source system for data, down from six.

Approach

The KPMG Cryptoasset Services practice brought together a broad range of specialized business and technical skills to build and deploy a robust AML transaction monitoring tool. Our team included crypto specialists, cybersecurity professionals, technology architects, data scientists, capital markets specialists, smart contract developers, regulatory compliance and financial crimes professionals, technology auditors, tax professionals, and accounting advisers.

Working closely with our client’s leadership, we performed a risk assessment of the company’s businesses to better understand their customers and activities to make recommendations of AML scenarios for our client’s review and possible implementation.

Then we set about streamlining immense volumes of transaction details and customer information, organizing it all onto a single cloud-based platform. Even though the data existed in multiple formats and sources, we were able to process it into the same format so that different types of transactions could be run through different scenarios and still yield accurate results.

Testing and tuning the system proved especially complex. Each scenario triggered an alert at a particular threshold level. We had to perform above- and below-the-line testing to reduce as many false positives as possible. It was critical to calibrate the best thresholds—neither too high nor too low—to be sure the alerts were true positives.

It was also critical to design and test a statistically relevant number of alerts—in this case, transactions between and among five traditional currencies and seven cryptocurrencies.

The crypto exchange’s new systems passed independent testing performed by an outside party. Completed in only nine months, the solution is scalable and reliable.

KPMG also built in advanced analytics capabilities to leverage artificial intelligence and machine learning technologies that can be used in the future.

KPMG insights

Considerable testing up front is the key to making AML alerts more accurate

We made thousands of small calibrations to help reduce the number of false positives so that the company could have high confidence in the results.

 

Cloud-based solutions are the wave of the future in the financial services sector

Though cloud adoption lags among traditional financial services companies—mostly due to security concerns—newer financial entities like cryptoasset exchanges and fintech firms are leading the way in offering leading-class security, compliance, technology, customer support, and design, among other services.

Even though the data to be monitored comes from multiple formats and sources, this client can now process it in one format. Different types of transactions can be run through a variety of scenarios and still yield accurate results.