Friday, 20 December 2013

Cash Management and sample ternd analysis


For any entity, cash management is very important. For a sovereign government it becomes further important as government is expected to honor its commitments without fail and incur several social and developmental expenses. Aim of public sector cash management is to forecast the availability of the total liquid cash resources at a point in time which is at the end of day, month, quarter, half year or year. To find out this simple figure a lot of data needs to be collected and then trend analysis needs to be done to find out reasonable requirement of funds. Any unusual or non-linear cash requirement, if estimated properly, will make the job of cash management simpler.
Since variable affecting cash flow are many and there are many factors influencing those variables, complex modern IT systems are available to produce cash flow projections and plans. However, before venturing into complex systems, developing countries must first use and let a system of collecting relevant data for cash management settle down and project cash requirements even without the use of complex software. It is possible to reasonably do the cash flow analysis and planning without an off-the-shelf IFMIS and debt management system with a cash management module.
Prudent cash management depends on sound revenue forecasting and tight project expenditure monitoring. An active analysis of historical data of expenditure alongwith assessment of requirements of future spikes and then a trend analysis of the same to cull out data in a scientific manner shall be good enough to get the projections within the acceptable limits. Since quality and reliability of data, in many developing countries may be an issue, hence, trend analysis may throw some data which will be off the mark. But following the model consistently and using corrections in the output, trends may be predicted to near perfection. Detailed analysis of the actual expenditures and its timing and revenues generated against their respective projections is a must. There are several components of expenditure, which are predictable in routine e.g. salary expenses, pension payments. In many countries, for such expenditures, it can be quite sufficient to develop a model for annual expenditure on the basis of monthly expenditure. Variations across and within months would not be significant in normal course. Any abnormal or having bearing on cash flow must necessarily be monitored and factored in while projecting cash flow.
There are several expenses incurred by government which are very volatile. It covers procurement and delivery of wide range of goods and services. In such cases, analysis need to done after obtaining finer details of expenses, may be even going to sub-chapter or line item level. This will eventually provide data to see where forecast errors have occurred and then corrective measures may be taken to avoid them in the future. Good source of data for elementary seasonal trend analysis is the historical expenditure database and annual budget line item appropriation. This may give an insight on efforts made by the line ministry in making an effort to produce accurate forecasts. For much of the analysis, the input data for the forecast will need to be fine. The annual budget by chapter and sub-chapter of the economic classification should provide ample data to project and figure out errors in projections. This data includes annual budget appropriations, revenue estimates, and debt servicing figures. This data may be taken in spreadsheet form and the historical database can be used to produce seasonal trend profiles for each sub-chapter (or item) which is then overlaid on the annual budget appropriation to produce an initial estimate of the expenditure or revenue for the desired timeframe. As the year progresses, there shall be a mechanism to collect further information and cash plans from the agencies and the budget department.
In an excel spreadsheet, data captured may be projected using different trend analysis techniques, depending upon the nature of data and variations therein. Data of revenue and non-tax revenue as well as expenditure of last five years have been tabulated and trend analysis used to project budget estimates of next two years for Government of India.
Table-1
amount in million INR

as per actual data


Budget
Actuals@


Estimates
for


2013-14*
April 2013


10563310
78980


Tax Revenue (Net)
8840780
31930


1722520
47050


16652970
1016640



Table-1 contains data of budget of 2013-14 and actual of Govt. of India in the month of April,2013. Table-2 contains budget data of 4 years prior to 2013-14.
Table-2
amount in million INR

Budget estimates (BE) as per actual data 

Budget Estimates

2009-10*
2010-11*
2011-12*
2012-13*
6095510
6822120
7898920
9356850
Tax Revenue (Net)
4975960
5340940
6644570
7710710
Non-Tax Revenue
1119550
1481180
1254350
1646140
Total Expenditure
9532310
11087490
12577290
14909250

Table-3 has this data and data of two future years derived through trend analysis. Chart-1 and Chart-2 has been prepared from these two tables. Linear trend is having very good level of reliability as value of R-squared is very close to 1.
Table-3
amount in million INR

Budget estimates(BE) as per actual data
BE as per trend 

Budget
Budget

Estimates
Estimates

2009-10*
2010-11*
2011-12*
2012-13*
2013-14
2014-15
6095510
6822120
7898920
9356850
10258555
11344637
Tax Revenue (Net)
4975960
5340940
6644570
7710710
8545015
9495803
Non-Tax Revenue
1119550
1481180
1254350
1646140
1713540
1848834
9532310
11087490
12577290
14909250
16431740
18193802

Chart-1
Chart of Linear trend of Budget Estimates: Linear projection
                                                                                                 (amount in million INR)
 


 Chart-2

Chart of Budget estimates of linear projected values
                                                                                                                     (amount in million INR)
 Table-4 and Table-5 has similar data for actual revenue and expenditure of the month of April during last 4 years.

Table-4

amount in Million INR

as per actual data


Actuals@
Actuals@
Actuals@
Actuals@


Apr-09
Apr-10
Apr-11
Apr-12

118460
129790
68800
191190

74620
100620
37740
152100

43840
29170
31060
39090

662170
672260
871300
877090

Table-5
amount in Million INR
as per actual data

Actuals@
Actuals@
Actuals@
Actuals@
Actuals@
Actuals@

Apr-09
Apr-10
Apr-11
Apr-12
Apr-13
Apr-14
118460
129790
68800
191190
166360
182080
74620
100620
37740
152100
133660
150616
43840
29170
31060
39090
32700
31464
662170
672260
871300
877090
981655
1066035


Chart-3
 Chart of Actual data with Polynomial trend analysis and projected values
(amount in million INR)


However, in the case of actual figures of April, variations are too significant; hence polynomial trend analysis has been used (Chart-3). It has given reasonably reliable projections with acceptable value of R-sqaured.
Chart-4
Chart of linear projected values of actuals
                                                                                                                       (amount in million INR)



No comments:

Post a Comment