Obtain the data you need to make the most informed decisions by accessing our extensive portfolio of information, analytics, and expertise. Sign in to the product or service center of your choice.
Funding valuation adjustment, or FVA, has long been the subject
of heated debate within the industry. Intended to ensure banks
account properly for the cost of funding unsecured derivatives
transactions, it is one of the concepts covered by XVA. However,
questions persist around how FVA should be applied and calculated.
As part of the Totem XVA service, we collect a range of submissions
from clients - including on FVA - which gives us unique insight
into patterns and trends. Banks are interested in comparing
submissions to their peers and so we have sought to review the
significance of various factors and how they affect submissions.
FVA is where we often see larger differences in pricing between
participants and so our analysis is focused specifically on this
concept.
In this analysis, the FVA submissions are reviewed in absolute
terms. The term 'above the consensus' refers to a larger negative
adjustment and 'below consensus' refers to a smaller negative
adjustment, as compared to the consensus data. One of the
significant variables in banks that submit to the XVA service is
the region where they are based. The question therefore arises:
does region have a significant impact on the adjustments that are
submitted?
Our analysis shows that there does appear to be some
relationship between region and USD adjustment. This relationship
is similar across the different scenarios, so we have chosen the
10Y swap as the sample scenario against the IG Sovereign
counterparty.
The USD swap shows that all North American banks submit 110% of
consensus or below whilst all APAC banks submit 90% of consensus or
above. The European banks show the most dispersion and submit data
across the range.
These findings indicate that region can play a factor in
submissions, although European banks appear to have the least
consistency. The picture is slightly different when looking at the
10Y EUR swap which shows a different spread. Although the European
banks are still spread across the range, there are a number of
banks from different regions submitting large adjustments. Both
APAC and North American banks show a much wider distribution,
varying from one of the smaller adjustments to one of the
largest.
The data indicates that for a EUR swap, the bank's domicile is
not a significant factor affecting its FVA, but there does appear
to be greater influence of domicile when looking at FVA on USD
trades. This is most apparent for North American and APAC banks.
The dispersion in European banks - when compared to other regions -
may be in part due to the differing local currencies. Not all
European banks have EUR as their local currency and this may cause
some dispersion in the results. A similar dispersion to the EUR
trades was observed in the GBP trade, where domicile was not a
significant factor.
Reviewing the internal relationship between currencies for banks
based in three regions and using the USD FVA adjustments as the
base, the average basis between GBP and USD trades is smallest and
the basis between AUD and USD is the largest. This demonstrates
that in all regions GBP FVA is, on average, closer to the USD FVA
than either AUD or EUR to USD. The AUD basis is positive,
signifying larger AUD FVA adjustments and the EUR basis is negative
representing a smaller EUR adjustment. The GBP basis is not
consistent across regions.
The domicile of a bank appears to affect the relationship
between currencies only with GBP to USD, which is likely influenced
by the relationship of region indicated in the USD trades
previously, as the North American banks' FVA was smaller on the USD
trades when compared to other regions, while there was no
relationship present in the GBP trade.
Of all of the factors we reviewed, region appears to have the
most significant effect on FVA with the other factors (credit
spread, rating and bank structure) having less impact. Data was
reviewed over a number of months and a variety of scenarios
analysed and no relationship was found between these three
additional factors. Therefore, it is highly possible that if there
is a relationship observed in these factors, then region may be
affecting them.
Posted 01 May 2018 by Vaughan Harding, Director, IHS Markit Totem
