# Machine learning critical for better SME credit scoring in trade finance
Author:  Pal Sinha, Barnali 
Author URL: https://financedigest.com/author/pal-sinha-barnali
Published: 2018-05-29
Category: FINANCE
Category URL: https://financedigest.com/category/finance
Meta Title: How AI and Data Are Transforming Trade Finance for SMEs
Meta Description: Learn how AI and broader data collection can revolutionize credit scoring for SMEs in trade finance, improving access and timeliness of credit decisions.
URL: https://financedigest.com/machine-learning-critical-for-better-sme-credit-scoring-in-trade-financehtml

![undefined](https://prod.superblogcdn.com/site_cuid_cm5qst7v3003gwirgwqtxn8i8/images/fd290518-8-1736842686320-compressed.jpg)

_AI and broader data collection can overcome the “trade finance gap” by improving credit scoring for SMEs in trade finance, says_ _Michael Boguslavsky, head of AI at Tradeteq and author of a white paper released today._

LONDON: Tradeteq, the trade asset distribution platform, has today released a white paper aimed at demonstrating how machine learning, combined with broader data collection, can improve access to trade finance for SMEs. Authored by Michael Boguslavsky, Tradeteq’s head of AI, and titled _Machine Learning Credit Analytics for Trade Finance_, the paper proposes a radical new approach to credit scoring that could particularly benefit SMEs in trade finance.

The paper states that traditional models – such as the Altman Z-score – use a “linear discriminant” analysis, which is based on several accounting indicators. While widely utilised, such scoring presents a number of issues for SMEs – including focusing on a small number of [accounting entries while ignoring valuable](https://www.financedigest.com/covid-19-has-made-your-customer-accounts-more-valuable.html "Covid-19 has made your customer accounts more valuable") non-accounting information. Such hard requirements make [credit scoring](https://www.financedigest.com/how-to-raise-my-credit-score-40-points-fast.html "How to raise my credit score 40 points fast") impossible for companies that miss even one entry. Being based on accounting data filed on an annual basis, traditional scoring also lacks timely information.

“Over the years there have been many attempts to improve traditional [credit](https://www.financedigest.com/12-things-that-dont-hurt-your-credit-score.html "12 Things That Don’t Hurt Your Credit Score") scoring,” says Michael Boguslavsky, “such as adding new financial ratios or replacing the Altman Z linear approach with other models. But they have never been very successful. What’s needed are models able to leverage non-homogenous data from [multiple](https://www.financedigest.com/how-to-navigate-multiple-data-privacy-regulatory-frameworks.html "How to navigate multiple data privacy regulatory frameworks") data sources – dramatically improving both quality and timeliness of credit event prediction.”

Boguslavsky’swhite paper argues that a good predictive credit model for [trade finance](https://www.financedigest.com/trade-finance-the-small-brush-to-paint-the-big-picture.html "Trade Finance: The Small Brush to Paint the Big Picture") lending should:

- accommodate varying data availability across companies to increase the depth of datasets,
- leverage a broad set of available and emerging data sources, including geographical data,
- utilise trade network data, including common clients, suppliers, or bank relationships, to spot irregularities and predict [credit risk](https://www.financedigest.com/audit-finds-holes-in-ecbs-management-of-bank-credit-risk.html "Audit finds holes in ECB’s management of bank credit risk").

It’s this approach that will allow for a broader understanding of SMEs’ credit risk, leading to fewer [loan rejections and improved credit](https://www.financedigest.com/4-ways-to-get-a-bad-credit-loan.html "4 Ways to Get a Bad Credit Loan") decisions, claimsBoguslavsky.

“The combination of [machine learning](https://www.financedigest.com/machine-learning-in-artificial-intelligence.html "Machine learning in artificial intelligence") techniques with deep and broad data coverage generates a neural network model that can outperform the traditional Altman Z-score and similar models even on pure registration data,” says Boguslavsky. “And this without using any accounting inputs – hence it’s [potentially revolutionary impact](https://www.financedigest.com/the-potential-impact-of-coronavirus-covid-19-for-riveting-tools-market-revenue-opportunity-forecast-and-value-chain-2021-2028.html "The Potential Impact Of Coronavirus (Covid-19) For Riveting Tools Market Revenue, Opportunity, Forecast And Value Chain 2021-2028") on SMEs seeking trade finance.”

Tradeteq’s [trade asset distribution platform](https://www.financedigest.com/exclusive-tp-icap-to-launch-crypto-trading-platform-with-fidelity-standard-chartered.html "Exclusive: TP ICAP to launch crypto trading platform with Fidelity, Standard Chartered") generates credit scoring in just such a way, with the aim of expanding the universe of trade finance investors by encouraging an “originate to distribute” model by trade finance banks. The company – officially launched in March 2018 – is now looking for partnerships and collaborations to work on transaction-level trade finance datasets, leveraging Tradeteq’s expertise in [deep](https://www.financedigest.com/low-calorie-toast-market-insights-deep-analysis-of-key-vendor-in-the-industry-2022-2031.html "Low-calorie Toast Market Insights, Deep Analysis of Key Vendor in the Industry 2022-2031") data analysis and the broad data sourced from partners to produce state-of-the-art credit analysis for the trade finance community. \[ENDS\]

**The white paper** _Machine Learning Credit Analytics for Trade Finance_ can be downloaded here.

_About Tradeteq:_

- Tradeteq provides a collaborative network for [trade finance investors](https://www.financedigest.com/how-to-begin-investor-trading-in-australia.html "How to begin investor trading in Australia") and originators to connect, interact, and transact. Tradeteq connects trade finance originators with funders and gives them the technology to interact and transact efficiently.
- The Tradeteq Marketplace delivers AI-powered credit analytics, reporting, investment, and operational solutions – transforming trade finance assets into transparent and scalable investments able to [attract institutional funding](https://www.financedigest.com/global-equity-funds-attract-inflows-for-second-week-in-a-row.html "Global equity funds attract inflows for second week in a row").
- The Tradeteq Marketplace helps trade financiers build an “originate-to-distribute” model – helping [banks overcome](https://www.financedigest.com/how-can-banks-overcome-the-tech-trust-gap.html "How can banks overcome the tech trust gap?") balance-sheet constraints within their lending portfolio by efficiently distributing trade finance assets to a broad investor base.
- After a soft launch in 2017, Tradeteq was officially launched in March 2018.


---
This blog is powered by Superblog. Visit https://superblog.ai to know more.
---