WITNO1050200
WITNO1050200
Witness Name: Ronald John Warmington
Statement No.:WITN01050200
Dated: 10 June 2024
POST OFFICE HORIZON IT INQUIRY
SECOND (SUPPLEMENTAL) WITNESS STATEMENT OF RONALD JOHN
WARMINGTON
I, Ronald John Warmington, will say as follows:
INTRODUCTION
1. I am currently a Director, and the Chairman, of Second Sight Investigations Ltd
(“SSIL’) and I have also served as a Director, since 1 April 2010, of its now largely
dormant predecessor company, Second Sight Support Services Ltd (“SSSSL’).
SSSSL was appointed, in June of 2012, by Post Office Limited (“POL”) to carry out
Horizon-related investigative work (SSIL was incorporated much later, on 31 May
2021). For ease, throughout this Witness Statement, I shall refer to SSSSL as Second
Sight.
2. I make this, my Second Statement, for the assistance of the Post Office Horizon
IT Inquiry (the “Inquiry”), as a supplement to my First Witness Statement, in order to
clarify, and to explain in a little further detail, some points/issues that seem to me to
merit clarification. These are: (1) the nature of suspense accounts; (2) my contention
Page 1 of 13
WITNO1050200
WITNO1050200
that Horizon was not, in respect of several products and processes, what can properly
be called a ‘double entry accounting system’; and (3) some further observations
regarding ATM-related shortfalls. I have not received legal assistance in the drafting
of this statement (Freeths Solicitors assisted me with my first witness statement).
3. SUSPENSE ACCOUNTS: I have heard the Inquiry many times seeking
clarification in respect of concerns, that were first expressed by Second Sight that have
been repeated by others and dismissed by Post Office, about Post Office’s
SUSPENSE ACCOUNTS.
4. I hope the following notes will assist the Inquiry. I realise that the Inquiry’s
Leading Counsel has already completely understood what I am about to describe.
5. First of all it is necessary to understand that there have been two types of
suspense accounts. The first type are no longer (since 2006) in existence. These were
the BRANCH SUSPENSE ACCOUNTS. They were available (to those SPMs that
knew of the functionality) to temporarily hold shortfalls (or, far less frequently, to hold
surpluses) that had been caused by errors the causes of which the SPMs had been
unable to establish and that they had therefore been unable to correct. The removal
of that functionality, as a component of the ‘Impact Project’, while at a stroke curing a
major headache for POL, triggered profound difficulties for some SPMs, leaving many
of them with no realistic alternative, when faced with shortfalls that they could neither
understand nor make good, but to falsify their branches’ books in order to ‘buy time’.
6. The second type of suspense account, about which there seems still to exist
some confusion, are POL’S OWN SUSPENSE ACCOUNTS. When I first asked POL
Page 2 of 13
WITNO1050200
WITNO1050200
about these (in pursuit of an answer to a question posed by Sir Anthony Hooper in a
Mediation Working Group meeting), the answer that came back, from Angela van den
Bogerd, was: “we don’t have any suspense accounts“. I explained that POL would
have suspense accounts, probably one for each client relationship.
7. I remember that it took an inordinate amount of time to get POL’s
representatives at the Mediation Scheme’s Working Group meetings to understand
this. It proved necessary to send very detailed emails before the penny eventually
dropped and POL even started to understand what I was talking about.
8. I formed an hypothesis, which I explained verbally and then spelled out in detail
in writing, that some amounts that ought to have been returned to SPMs might have
been credited to POL’s own Profit and Loss account.
9. It might be best if I give an example of the sort of situation that might lead to that
having happened. So, by way of example:
a. A customer goes into a branch and spends £1,000 on Premium Bonds
using a debit card to pay for them. If there is, at the moment of the
transaction, a transient communications blip, the customer could walk
out with those Premium Bonds but without actually having been charged
for them because one side of the transaction went through (the sale of
the Premium Bonds went into Horizon) but the other side (the debit card
transaction that was meant to have been processed, through the LiNK
system, to debit the customer's bank account) had (at that stage
unbeknown to the customer or to the branch) failed. That would be an
Page 3 of 13
WITNO1050200
WITNO1050200
example of what we (in Second Sight) referred to (and POL in due
course also now refers to) as a ‘one-sided transaction’.
. That scenario would mean that the customer would then be holding real
Premium Bonds for which he or she had never paid. There was in fact a
real example of this happening, back in 2007, when an honest and
diligent customer wrote to The Sunday Times (her letter having been
published in its ‘Question of Money’ column on January 28th, 2007)
about a similar type of windfall, albeit that one was probably caused by
the customer’s cheque having been mutilated in one of POL’s cheque
clearing machines. See my email, dated 13% February 2013, that
mentions this case:
. It is clear to me (one might suggest that this is inescapable logic) that,
when a customer gets something for nothing, he or she might give way
to the temptation to keep quiet about it - or maybe even he or she would
not even have noticed the error in the first place, particularly if the
amounts were - to them - relatively small. If my logic is correct, more
customers will notice - and then complain about - having received
nothing for something than there will be customers who report (as did
the lady who wrote to The Sunday Times) having received something for
nothing.
. Clearly, when a customer gets something for nothing, that will (unless
the customer's bank, or POL, absorbs the loss) create a shortfall in the
branch, because the customer has something, and POL hasn't received
the payment for it. POL, centrally, will be the first to find that out... most
Page 4 of 13
WITNO1050200
WITNO1050200
likely the next day... and it will then send a Transaction Correction (a
‘TC’) ‘Invoice’ to the branch in order to recover the missing money.
. When that TC is accepted by the branch, it will generate, in the example
we are using here, a £1,000 shortfall in that branch's books.
That shortfall will only be zeroed out in the branch’s books IF the
customer lets his bank know that he’s received something for nothing...
and IF that bank debits his account and sends the £1,000 to POL... AND
AS LONG AS POL THEN ASSOCIATES THAT INCOMING £1,000
WITH THE RIGHT BRANCH IN ORDER TO CORRECT THE
SITUATION... by sending a ‘Credit Note’ TC to that branch.
. So... in summary, POL’s (central) books in Chesterfield will have briefly
shown a £1,000 shortfall, when the customer's bank failed to send that
amount to POL. And POL will have immediately then onward charged its
£1,000 shortfall to the branch by sending it a TC ‘Invoice’, thereby
transferring its shortfall back to the branch.
It should now be clear that the branch is not going to get its SPM’s
£1,000 back until (and unless) the ultimate customer who has benefitted
from the windfall:
i. notices that he’s not been charged (this might be months later)...
and
ii, then notifies/confesses to his bank that he has received
something for nothing... and then
iii. that bank client of POL then contacts POL to tell them about it...
and
iv. then it pays over to POL the £1,000... and then
Page 5 of 13
WITNO1050200
WITNO1050200
v. POL correctly (and this might be many months later, during which
time the SPM has had to fund the shortfall) matches up its
incoming £1,000 with the original error... and finally if...
vi. POL then sends out a Credit TC to the branch to correct (i.e. to
remove) the shortfall that it had generated for that branch when it
sent out the TC Invoice.
It ought to be obvious that any of steps i. to vi. can be delayed, or they can fail
to happen at all, in which case the end result will be that the SPM will have paid
for the customer's Premium Bonds and either the customer, or his bank, or POL
will have benefitted by keeping the £1,000. Of those three possibilities, the least
likely, in my opinion, would be for the customer's bank to have benefitted.
In that example scenario, both the customer's bank, and POL, would have put
the £1,000 difference into a ‘client suspense account’, in their own company’s
books, where it would sit, as an ‘unreconciled item’, until the root cause of that
difference - between the two companies’ books - was identified and corrected.
10. If the process fails at point v. above (i.e. having received the £1,000 from the
customer's bank, POL’s people in Chesterfield have failed to match it to the one-sided
transaction and therefore to the branch that suffered the shortfall) then that amount
will remain sitting in POL’s client suspense account (i.e. the one for that customer's
bank) for three years, and then, having been subsumed in the residual balance in that
client suspense account, it will be credited to POL’s P&L Account, thereby increasing
its profits.
Page 6 of 13
WITNO1050200
WITNO1050200
11. It follows that any amount that remains sitting, as a component of the balances
in any of POL’s client suspense accounts, will always, in truth. properly be owned by,
and therefore be properly returnable to, one of only four potential owners. Those four
are:
1) POL itself
2) One of POL’s CUSTOMERS
3) One of POL’s CLIENTS (such as HMRC, Camelot, Royal
Mail, Bank of Ireland, Santander, the DVLA, NS&l, etc,
etc.)
4) An SPM
12. The bottom line on all of that is that my colleagues and I formed an hypothesis,
that we wished to test (and hopefully to disprove), that some of the
UNRECONCILED net credits that POL had every year (after three years)
credited to its own P&L Account might have included amounts that, had POL
been able to establish to whom they truly belonged, ought to have been
returned to SPMs. We later learned, from POL when it wrote to us about its
Suspense Accounts, that it had released, during the year 2010/2011, gross
unresolved credits that aggregated to £612,000.
13.Had Second Sight been allowed to examine POL’s suspense accounts, we
would have:
a. examined to whom entries that POL HAD been able to clear had been
returned to (i.e. over time, what percentage of cleared items were found
to be due to each of the four parties listed above);
Page 7 of 13
WITNO1050200
WITNO1050200
b. been able to estimate, based on the percentages derived in step (a),
broadly how much of the remaining, UNRECONCILED, balances might,
had they been bottomed out, have been found to be due back to SPMs.
If it turned out that that figure was zero, then we would have disproved
our hypothesis;
c. focussed on the very large balances in the Santander, ATM, and
Camelot suspense accounts, in order to gauge the real-life impact of
those large transient balances on SPMs. For example. If component
entries in POL’s ATM-related suspense account meant that SPMs had
needed to somehow fund their apparent shortfalls during the weeks - or
months - that it took for the related suspense account entries to be
investigated and resolved, then had that caused such grief that some of
them had, post-2006 (i.e. after the local branch suspense accounts were
removed as part of the ‘Impact’ project), decided to falsify their branches’
declared cash balances in order to ‘buy time’ to try to find out what had
caused their shortages?; and...
d. taken the examples that had surfaced during the Mediation Scheme (e.g.
in case M082, Mr Peter HOLLOWAY... and the double-paid customer
Income Tax bill), and chased those individual examples through POL’s
suspense accounts to test out (and hopefully to disprove) our
hypothesis.
14. I have seen no evidence that those lines of enquiry were ever pursued by Post
Office or by any of the accountants that it engaged to provide reassurance that Second
Page 8 of 13
WITNO1050200
WITNO1050200
Sight’s suspicions, as to the possibility that Post Office had taken money that ought to
have been returned to SPMs, were groundless.
15. Further to the above, it is my belief that it wasn’t - generally - ‘bugs’ that will have
caused the majority of the alleged disappearances of money, but rather it will have
been, as POL itself has asserted: “errors made at the counter’. I also contend that
many of those errors will have been caused by preventable system and process
design flaws.
16. I have also suggested (as, later, did Deloittes) the possibility that some of the
shortfalls, for which SPMs have been held accountable, might have been fraudulently
generated by ‘Superusers’ in Fujitsu sending shortfalls to victim branches and sending
the aggregated credits as surpluses to branch SPMs with whom the perpetrators were
colluding. The terrible consequence, if that really has happened (although there exists
no evidence, of which I am aware, that this has happened), is that some of the victim
SPMs would, if they had then hidden the (unaffordable) shortfalls, not only have
committed the criminal offence of False Accounting, but they would have also, in
effect, unwittingly colluded with the perpetrator(s) by hiding thefts where they
themselves (i.e. the SPMs) were the victims.
17. CAN HORIZON PROPERLY BE CLAIMED TO BE A ‘DOUBLE-ENTRY’
SYSTEM?: I referred, in paragraphs 75 to 82 of my First Witness Statement, to Post
Office’s handling of Foreign Currency (“FX”) transactions. We in Second Sight also
referred, in many of our CRRs and in our 9" April 2015 Briefing Report - Part Two
[POL00029849], to the risks that processing Lottery, ATM and FX transaction posed
Page 9 of 13
WITNO1050200
WITNO1050200
for SPMs. It is relevant to note here that Detica also drew attention, in its October 2013
‘Non Conformance’ Report, to the materially increased risk of an SPM failing an audit
where those products/services were being delivered. I should add that we, in Second
Sight, had no knowledge of Detica’s work, or of its report, until years later, well after
the GLO had been concluded.
18. I submit here that one of the root causes of those materially increased risks was
that those transactions were not processed TRANSACTIONALLY through the Horizon
system. Instead, Lottery and ATM transactions were only entered into Horizon /N
AGGREGATE ON A DAILY BASIS, and FX transactions were only processed, again
IN AGGREGATE, ONCE A WEEK. That sort of bulk processing of blocks of
transactions cannot properly be described as “double-entry bookkeeping in the
Horizon system”. I suggest that assurances that have for decades been given - and
that have remained unchallenged - as to Horizon being “a double-entry bookkeeping
system” may, therefore, be unfounded. What I am asserting here might also have
some relevance to the categorisation of some shortfall cases as “Horizon cases”
whereas others have been categorised as “non-Horizon cases”. The significance of
such categorisation is, it seems, that “Horizon cases” are now regarded as being:
inherently suspect (because they were exposed to known, or as-yet-unknown, Horizon
‘bugs’), whereas “non-Horizon cases” seem to have been regarded as somehow
inherently safe. I do not believe such thinking to be sound. Indeed, I cannot see why
losses that were caused by, or that appear to have been caused by, Lottery, ATM or
FX transactions (or by any other products or services that were not
TRANSACTIONALLY processed through Horizon), should be regarded as “Horizon
cases” at all... given that those transactions were never processed (individually)
Page 10 of 13
WITNO1050200
WITNO1050200
through Horizon. I also suggest that, although it is correct to conclude that shortfalls
that arose from Lottery, ATM and FX processing could not (of course) have been
caused by bugs in the Horizon software, it is feasible that they could have been caused
by bugs in the software deployed in the various front-end transaction processing
systems that POL had ‘bolted onto’ Horizon (this was also referred to by Detica in
paragraph 7.2.2. of its October 2013 Report [POL00029677] where they reported:
“difficulty in reconciling information from multiple transaction systems both in terms of
timeliness, structure and access’).
19. In summary, it seems to have been accepted that those prosecutions that have
relied upon Horizon data are all unsafe, whereas those that were based upon data
derived from other sources (including figures supplied by POL’s clients such as the
DWP, Bank of Ireland, Santander, Royal Mail, etc.) were, or were more likely to have
been, safe. I suggest that they were/are not safe, and that the acceptance that they
were all safe is not only founded on an unproven premise, but on a premise that has
never, to my knowledge, been tested.
20. ATM SHORTFALLS: I would like to elaborate a little on the ATM-related risks
mentioned in paragraph 17. The point that I feel needs to be stressed is that the sheer
magnitude of many ATM-related shortfalls meant that they were highly material to the
impacted SPMs. We drew attention to this in paragraph 7.3 of our Briefing Report -
Part Two, where we said that more than 20% of the Applications to the Mediation
Scheme had referred to ATM shortfalls and that thirteen of those ATM-related
shortages exceeded £20,000, six exceeded £60,000 and one exceeded £80,000. As
far as I am aware, very few witnesses, aside from Mr Alasdair Cameron, have even
Page 11 of 13
WITNO1050200
WITNO1050200
recognised, let alone acknowledged in their testimony, that shortfalls that were clearly
“immaterial to POL’s staff, and to its external auditors, in the context of POL’s books,
would have been enormous (indeed life-changing) in the books of a branch.
21. Wehad also referred, in paragraph 2.18 of our Briefing Report - Part Two, to the
fact that the unreconciled Bank of Ireland ATM-related Suspense Account balance, at
the 2014 financial year-end, was £96 million. As far as I am now able to recall, Post
Office asserted that no entry in that Suspense Account was more than six months old
but that seemingly reassuring statement ignored the likelihood that some very large
shortfalls will have remained uncorrected in branch accounts for some months... so
that the impacted SPMs will have needed, in the meantime, to make good those
shortfalls. Where they were unable to make good what would, in some cases, be
shortfalls of tens of thousands of pounds, they may have fallen victim to the temptation
to falsify their branches’ accounts while ‘waiting for something to turn up’.
Statement of Truth
I believe the content of this statement to be true.
Dated:___ 10" June 2024
Page 12 of 13
Index to the Second Witness Statement of Ronald John Warmington
WITNO1050200
WITNO1050200
No
URN
Document Description
Control Number
Email from Ron Warmington to Jo
Hamilton cc Alan Bates, Kay
(Linnell) and lan Henderson re
Missing Cheques
POL00029677
Draft Report on Fraud and Non-
conformance in the Post Office;
Challenges and
Recommendations
POL-0026159
Page 13 of 13