LCAS0000974
LCAS0000974
Date: 25 Suly 2006
Your ref: $JD3/FACI/348035.134
Our ref: MDT.113969
Please ask for: Mark Turner __.
Direct dial: i
Direct fax :
E-mail:
Bond Pearce
Solicitors
DX 8251
PLYMOUTH
Dear Sirs
Post Office Limited -v- Mr L Castleton
Following disclosure, our client has been reviewing the transaction logs disclosed by your client. We
understand from paragraph 2(c) that these documents constitute a full audit trail. By that, we understand that
these logs are intended to show an item-by-item statement of all transactions carried out at Marine Drive Post
Office for the particular days shown in the logs.
Mr Castleton has carried out a manual reconciliation of all transactions recorded in the log for Week 42. Since
this is a time-consuming exercise, we have for the time being limited the analysis to that week, although in
light of what we say below we envisage that it will be necessary for each of the following weeks in which it is
alleged shortfalls arose. Week 42 was chosen for this initial analysis as it was the first week in which an
apparent discrepancy arose and therefore the starting figures going into Week 42 are undisputed and provide
an uncontentious benchmark against which to judge the movements during Week 42.
Mr Castleton’s analysis of the transaction logs for Week 42, and a comparison with the cash accounts for that
week appears to show a number of inconsistencies. According to the final cash account generated at Marine
Drive Post Office for Week 42, total receipts of £176,291.15 were paid in during the course of Week 42. The
end figure carried over from Week 41 was £54,170.49. As a consequence, it can be seen that, apparently,
£122,120.66 was received during Week 42. However, a manual reconciliation of the figures contained in the
transaction log for this week gives a figure for monies received of £125,013.90, an apparent difference of
£2,893.24.
Similarly, an analysis of the figures relating. to payments (i.e, monies leaving Marine Drive Post Office during
Week 42) also shows a discrepancy. The figure for stock carried over from Week 41 was £92,374.74. If this is
deducted from £176,291.15, this leaves a figure of £83,916.41 in respect of payments made. Again, however,
a manual reconciliation of all of the transactions showing in the transaction log gives a figure of £83,707.07
This difference is not allocated in the Week 42 cash accounts.
It would further appear that there is £1,706.64 in unallocated cash shown as being paid out of Marine Drive
Post Office. The computer system then seems during the end-of-week balancing to have offset this
unallocated cash against the £2,893.24 figure, along with some further giro bank transactions, and produced
an apparent cash shortfall figure of £1,103.13. This is of course the figure that our client was required to make
good.
Quay
2 © Quay Street
DX 14352 MCR-I © Email
at
Ry. Sproston # S.Room + A.Curwen + Ry.tyer # H.Burns # SRSutton Associates: LF Swerling # AD. Owens « M. Molloy # F Sampson Consultants: I, Rowe # MT. Horwich
‘als frm ts ruled Oy the Law Society
Also at London nN Pi
(GAMARKT\ABBEV\CASTLETONWS070§ LETTER TO BOND FARCE
LCAS0000974
LCAS0000974
Our client has not been able to reconstruct precisely how Horizon arrived at the figure of £1,103.13. He
surmises that the system has sought to offset the £1,706.64 and other unallocated giro bank transactions
against the £2,893.24 figure set out above. We would invite your client to undertake its own analysis of the
transaction logs and provide an explanation for how that figure has been derived, in order that this discrepancy
can be explained and any possible flaw in our client’s methodology in carrying out the manual reconciliation
brought to light as quickly as possible.
‘This represents an analysis of the figures at their most basic level. Mr Castleton’s findings seem on the face of
it to be entirely consistent with his assertion since this dispute first arose, namely that the figures provided by
the Horizon computer system are incorrect. At the very least, it raises question marks about how a system that
relies in essence on double-entry book-keeping can give rise to such discrepancies.
Our client then proceeded to review the Week 42 transaction log in more detail. It would appear that a number
of the transactions recorded are incomplete — that is to say, only one half of a given transaction has been
recorded by Horizon, with no corresponding entry. Ordinarily, one would expect to see two “halves” of a
transaction. For example, a sale of £10 worth of stamps would show a £10 reduction in stock and a
corresponding £10 increase in cash. In a number of cases, the log shows apparent transactions with only one
“half” having been recorded by Horizon, and no corresponding other “half” to balance it.
In practice, this cannot happen. We understand from Mr Castleton that the Horizon system is configured in
such a way that in order to complete a transaction on screen a valid method of payment must be selected. As
such, the system would not process a transaction unless there had been a corresponding second “half” (or, in
the case of payments, a corresponding first “half”) to it.
The obvious inference to be drawn is that, if the system cannot be operated to only create half of a transaction,
Horizon is not recording the apparently missing “half”, Again, this is consistent with our client’s supposition.
that transactions have not been fully recorded (or recorded at all) as a result of problems in communication
between the two computer nodes used at Marine Drive Post Office.
Our client’s firm belief is that if half of a transaction were to be lost, the effect would only show in relation to
cash because Horizon looks for everything and the remainder is cash. If the first “half” of a transaction were
lost, it would show as having no effect on stock but would only be noticeable in terms of movement in the
cash figure. This also rather begs the question of whether whole transactions can be “lost” (i.e. not recorded
by Horizon) if it is possible to “lose” halves of transactions.
We appreciate that this is, of course, something that is going to have to be addressed in expert evidence. It will
also be necessary to undertake the same reconciliation in relation to the other weeks in which losses
apparently arose. [lowever, at this stage, we would invite you to conduct your own analysis of the figures for
Week 42 and consider these anomalies with your client, and then provide an explanation for them.
ROWE COHEN
GAMARKTABBEVCASTLETON(50705 LETTER TO BOND PEARCE