Traditional valuation models are under strain as data becomes both more abundant and less reliable. Danny Dieleman explains how shifting market conditions, illiquidity, and model risk are challenging pricing accuracy, forcing banks to rethink governance,
Valuation frameworks long considered
reliable in financial markets are coming under increasing pressure as banks
confront a paradox of modern finance — more data than ever, but less certainty
about how to use it.
According to Danny Dieleman, Wholesale
Banking Director, Capital Treasury at ING, the growing reliance on real time
data has introduced new vulnerabilities into core banking processes,
particularly in pricing, risk management, and capital calculations.
“There is probably more data
available than ever in human history,” he said, noting both the scale and speed
at which information is now generated and consumed.
However, this abundance does not
necessarily translate into reliability. Dieleman warned that increased
dependence on data exposes institutions to risks around data quality and
availability, particularly when models are used to value complex financial instruments.
“All models are in essence a
simplification of reality and should therefore be treated with care,” he said,
highlighting the limitations that become more pronounced in uncertain or
rapidly changing market conditions.
One of the most significant
challenges arises in markets where data is either scarce or becomes unreliable
during periods of stress.
Even traditionally liquid markets can
quickly deteriorate, leaving banks struggling to determine accurate valuations.
“Instruments that can normally be
priced by a simple look at a Bloomberg terminal can in the blink of an eye
become difficult to price,” Dieleman said.
This creates a dual risk. Pricing may
deviate significantly from underlying value, or, even if accurate, firms may
struggle to justify their valuations to auditors and regulators. In both cases,
the credibility of valuation frameworks is called into question.
Back testing, a cornerstone of model
validation, is also facing new limitations. While it remains a critical tool
for assessing model performance, its effectiveness depends on the availability
of representative market data.
Where observable prices are limited
or distorted, the insights derived from back testing become less reliable.
“As a consequence, it cannot be ruled
out that there is more model risk,” Dieleman said, pointing to the difficulty
of validating both model assumptions and calibrations in such environments.
