POL00141432
POL00141432
Post Office Branch Audit Trend Analysis YTD QI] 2012/13
Purpose of this paper
¢ To summarise all audit activity throughout Periods 1-3 2012/13
¢ To provide an overview of the findings from this activity
© To identify steps to mitigate the risk of potential financial loss to POL.
¢ To identify lessons learnt from audit activity as a way of informing future post office
network audit approach.
Audit Activity — Periods 1-3 2012/13
603 post office planned branch financial audits were undertaken from April 2012 to June
2012 as detailed in the table below. Explanation of audit type is at appendix I
Audit No Loss I Loss I Loss Loss Loss I Loss No Gain Type of No of
Type of under I £1k- I £5k- I £10k- I £20k- I over I discre I s I Discrepancy I Agents
Audi I £1k* I £5k £10k £20k £50k I £50k I pancy ee Suspende
it d
5 81 1 77 3 Cash
10 47 2 44 I__I Trans error 5
100 153 90 ia 5 2 i 1 43° I Cash & 7
stock
115 10 6 2 2 Cash
140 10 3 1 2 Cash
150 21 13 1 7 Cash 1
160 24 12 1 1 10 I Cash
161 10 8 2__I Cash
175 12 7 1 4 Cash
200 72 28 16 3 4 3 3 15 I Cash & 19
stock
250 5 3 1 1 Cash & 1
stock
300 153 89 3 1 1 9 50 I Cash & 2
stock
350 5 3 1 1 Cash, stock, 2
TC
Total 603 I 259 39 13 au a 3 133 142 37
% 43% I 6% 2% 1% 1% I <1% I 22% I 24% 6%
* 89% of these losses were between £1-500 with 66% being less than £200. ** 75% of these gains were less than £100
Performance against plan Q1 2012/13
Throughout Q] performance has been reasonably consistent with the 12/13 audit activity plan
but with a steady rise month on month in Special Audit requests.
Summary of Audit Findings YTD @ Quarter 1 2012/13
e 54% of all audit types undertaken identified a loss
¢ 24% generated a gain
© 22% had no discrepancy.
The two main areas of discrepancy were:
* Cash only at 30%
« A combination of cash & stock at 28%.
In Strictest Confidence 1
POL-0142817
Unprocessed or incorrectly processed transaction corrections accounted for 8%
of discrepancies.
¢ 4% included discrepancies in cheques on hand
Please Note:
It is important to note that that any discrepancy discovered on audit isn’t necessarily the
actual loss or gain declared by the agent at the branch trading (done monthly) as the audit
verifies the cash and stock on hand at the time of the audit which isn’t always undertaken on
the Thursday following that branch’s trading period date. Some errors that had occurred
within the branch trading period (4 weeks) could have been corrected in advance of the
branch trading declaration being made.
37 Agents were precautionary suspended as a result of audit acti
undertaken. Of these:
¢ 19 were from 200 Special audits (26% of this audit type) This remains the most effective
type of audit activity for identifying loss, with a YTD effectiveness rate of 51% of all
suspensions resulting from Type 200 audits
e 7 -suspensions (4.5% of this type) resulted from audits identified by the BPP in QI witha
YTD effectiveness rate of 19% of all suspensions resulting from Type 100 audits
5 (13.5%) from type 10 ( Robbery/Burglary)
1 (3%) from type 250 (FAA that changes to Tier 2 on site)
1 type 150 (Random audits) produced 3% of suspensions from this audit type.
2 (5%) from type 300 (Post Transfer at 6-9 months)
2 (5%) from type 350 (audit following cash check).
ity; this is 6% of audits
eeceee
Action to mitigate the risk of potential financial loss to POL
Branch Performance Profile
To further mitigate the financial risk to POL the new Branch Performance Profile (BPP)
has been deployed since Period 10 2011/12. The BPP introduced in Q4 2011/12 has continued
to be used in Q] to focus on high risk branches, and has driven 153 audits this quarter
resulting in a reasonable level of success in PI with c£73k found at audit resulting in 5
suspensions. In Period 2 this success was maintained, with c£66k found at audit resulting in 2
suspensions. However, during Period 3 the BPP was less successful, with only c£16k found at
audit and no resulting suspensions.
In Period 3 only 15 of the 51 audits delivered from the risk model identified a loss and 33 of
the losses identified were for less than £1k, with 23 of these for less than £500.
This reveals a trend of decline in terms of successful identification of loss through these data
streams and at the end of QI no longer reflects the positive performance over Q4 2011/12 in
the initial 3 month pilot period when 15 suspensions (8% of type) identified c£278k of loss.
The overall position for QI is 7 suspensions with an YTD effectiveness rate of 19% of all
suspensions resulting from Type 100 risk audits.
The decline in effectiveness of this risk model is a concern and brings into question the
validity and accuracy of some of the data feeds. The BPP is owned by the
Security Team — a review of the model is already underway to ensure we are focussing audit
field resource on the highest risk branches.
Type 200 Special Audit — Special Request usually from P&BA, the Fraud Forum or the
Security team in response to concerns ~ includes check of all cash, stock, vouchers relating
back to the previous Branch Trading Statement (BTS). This audit type delivered 19 of the 37
suspensions in Q1 (51%). Losses identified total c£466.7k
In Strictest Confidence 2
POL00141432
POL00141432
POL-0142817
POL00141432
POL00141432
Lessons Learnt
¢ The BPP model was not as effective in period 3 as it had been since its introduction in
period 10, 2011/12.
Action taken — review is underway to understand the reason(s) for the dip in effectiveness
and to identify corrective action.
¢ The special request audit (audit type 200) continues to be the most effective audit
activity.
Action taken — the Fraud Forum reviewed in Q] the end to end process for initiating a
Special Audit with a view to ensuring the robustness of the process in light of a number of
experienced people within P&BA moving to the Network Transformation Programme.
© Losses totalling c£748,910k were recorded at audit in Q1 2012/13 and outstanding debt at
audit (debt already in the system) was c£364,181k. Discrepancies in branch are typically as
a result of a cash error over the counter (eg not enough cash or too much cash given to or
taken from the customer per transaction); incorrect keying in of an amount to the Horizon
system (eg entering a £40 manual cash deposit as £400 and only taking £40 in cash would
result in a £360 loss on balancing); incorrect processing of documents (eg putting cheques
accepted into the mail stream rather than despatching to the processing centre would result
in a transaction error notice being received to reverse the value of the cheques from the
Horizon system resulting in a loss in branch); fraud or theft within the branch.
A couple of existing agents have claimed that their discrepancies are as a result of the
Horizon system.
Action taken — the process we follow to investigate claims from agents that the Horizon
system is generating discrepancies has been reviewed. The refreshed process is detailed at
appendix 2.
Overall Conclusion
The analysis of the audit findings for Quarter 1 2012/13 has shown some consistency in the
number of branches recording a loss at audit throughout the year at c54%.
The percentage of agents precautionary suspended as a result of their audit findings was 6% -
this is consistent with Q1 11/12. Asa percentage of the total number of audits delivered, the
highest number of suspensions came from Special Audits 200 (19 of the 37 suspensions) and
the Branch Performance Profile 100 (7 of the 37 suspensions).
The trends in the audit findings will continue to be analysed on a monthly basis and ongoing
learnings will continue to be taken from the audit activity throughout the year and where
appropriate improvements made within the year.
Angela Van Den Bogerd
Head of Network Services
7" Sept 2012
Appendix 1
Audit Types 2012/13
There have been a number of additions and other changes to the audit codes for this year, to
reflect different drivers and priorities during the Network Transformation Programme (NTP).
The full and current list is now:
[ Audit Code
In Strictest Confidence 3
POL-0142817
POL00141432
POL00141432
5 BAU Transfer from Agent to newly appointed Agent
10 Branch Transfer to/from Temp Agent
20 Robbery/Burglary
30 BAU Branch Closure Audit
35 Branch Closure Audit following Suspension at Audit
40 Conversion to Mains (existing Agent)
50 Conversion to PO Local ( existing Agent)
100 FAA Branch Performance Profile
115 Induction programme support
150 Random Audit
140 WHS(BAU)
141 WHS Audits (FAA Branch Performance Profile)
160 Crown Branch Audits(BAU)
161 Crown Branch audits(FAA Branch Performance Profile)
175 Branch Audit within 6 months, reinstated Agent following suspension at previous Audit
200 Special Request
250 FAA that changes to Tier 2 on site
300 6-9 month Post Transfer Financial Audit
350 Escalation following a cash check
OMPLIANCE
400 Compliance Audit with Financial Audit ( BPP, Random or Special request)
450 6-9 month Post Transfer Compliance Audit with Financial Audit
500 Compliance Only Audit - by appointment
550 Follow Up Audit (BST Compliance Issue)
600 Cash Centre Audit
Appendix 2
In Strictest Confidence 4
POL-0142817
Horizon Challenge Process
. Wednesday, August 29, 2012
secu Dee
ger
‘This isthe log of all requests
held by Team Leader of
Branch Support
Contract Advisor
This is the Horizon
Challenge Log sheet which
Is held on EFC but which is
‘also accessed and
summarised by Team
Leader of Branch Support
Contract
I Advisor
Enter onto
{ SharePoint
\ “suney I
is now the lead
Stage 1 & 2 of Horizon
Challenge Case Summary
Establish Branch
Set up
Establish what
prompted this
‘contact
Stage 3 of Horizon
Challenge Case Summary
“coun ads
rig
crgusence
reieenens
“arange FEA vats
lane any omer sang
ou
‘apes Ss 4 og
‘Yes
X
AB agent
satisfied with > No ons
‘Advise unable to No Request
progress further investigation via
untiican do so Duty Manager
‘Capture on Stage CA shares
p 5&7 Hotizon ‘outcome and
Challenge Case ‘actions with agent
immary at meet
AS
\ “Isa review ~ Yes
\ date in place. -
In Strictest Confidence
‘Supported by experts it
required ie. Fujtsu,
\ Meeting to be recorded or
minutes. NESP or fiend
‘can attend
POL00141432
POL00141432
POL-0142817