Why accountants should master information systems

When I refer to accountants in this context, it does not refer just to 'financial accountants' but also to 'management accountants' and corporate financial analysts, who handle internal corporate data. Of course, external auditors etc are anyways dutybound to evaluate client information systems to decide the extent of reliance on records/nature of assurance they can give, so I do not include them in this. Before understanding the ISACA study material of the ICAI, I felt that information systems was a geeky subject best left to engineers, but then I realized its real importance later. After all, accounting is an information system by itself, and therefore the IT approach to IS is just one aspect(but very helpful too, though it focusses more on automation and lesser on controls which accountants are expected to design).

1. System framework is quite useful for analyzing organization phenomenon. Skills using data flow diagrams etc are quite helpful while designing, auditing and improving systems.
2. To preserve their identity, autonomy and core competence, organizations are a closed system in many aspects. So are accounting systems-closed when it comes to basic principles, but quite open in the adaptation/interpretation of them.
3. IPOS cycle(Input process storage output) misses out in fact that feedback loops lead to output affecting input/virtuous cycles etc. This is true for accounting as well(which has evolved as per the way various users accounted their transactions).
4. Once we define the system boundary, we are much closer to defining the problem and our focus.
5. Example of buffer is cache memory, inventory, queues etc. This concept helps while scheduling work during peak seasons.
6. Way to decouple databases(lessen need for communication) is having common database standards, which is exactly what is done by many organizations(auto generated invoices between partners sharing same standards).
7. ERP is a more subtle form of putting in best practices, which can improve controls and performance substantially.
8. Post Implementation reviews like in PDCA cycle must to see how to get better next time, especially in the project management scene of IT. Those lessons useful for audit/control.
9. Outsourcing needs among other things, source code escrow, risk/controls and metrics for SLA. These should involve the finance function as well.
10. User signoffs not just CYA(cover your a**), also way to ensure buy in/proper needs built in etc

There can be tomes but this is all for now. 

Tips on how to analyze economic time series data

Be it stock prices, index valuations, budgetary spending, deficits, exchange rates; virtually any significant economic data is analyzed using time series data. In these days of Excel 2007, it is a great temptation to just plug the numbers in, select a graph and use it for whatever purpose one wants. But such a simplistic approach would be doing an injustice to the analysis, and also would reduce the chance that the data is analyzed, understood and appreciated in the way it should be. So what to do? Based on my not insignificant experience of this, I add some guidelines below. I don't claim this to be an exhaustive list, and comments are welcome

• Adjust for Inflation:-Specially for monetary values and interest rates, this is crucial. And sometimes it gives very interesting insights. For example, the real interest rates in India have been negative for quite some time, and that makes us appreciate why the RBI is chary about reducing it.
• Try per capital/per unit numbers:-It allows for adjustments for population/volume growth, and yields interesting results. Thanks to its population, India lags behind on such data
• Put the numbers in context:-Expressing as market share, budget share, relative data etc puts the numbers in context, like in per capita.
• Use Indexing if possible:-That makes the growth trend easier to understand
• Use secondary axis if useful:-For example, you want to show that FDI inflows fell when the exchange rate appreciated. Instead of using 2 graphs, you can use 'Plot on Secondary Axis' option in Excel, and plot them on the same graph
• Try to cover an entire economic cycle:-Otherwise, the result may just be because we are in a boom or a burst. This is true especially for stock price/valuation data
• More the better:-We often tend to forget history, and that is a reason why bubbles repeat, whether it be across countries or across time periods. Hence, a longer data set with more sources, helps to triangulate trends and have a broader perspective. 
• But be careful of underlying structural changes:-With Excel 2007, one can use callout and other options to depict the reasons for inflection points. Use those with care where warranted. For example, the 1991 liberalization would have lead to inflection point in most Indian data.
• Strike a balance between aggregates and subclasses:-Given the high Gini Index/Digital divide etc, one should be careful about what the data signifies for India as a whole. Often, the divide between states/castes/gender/urban-rural population parameters, would be large enough to vitiate any extrapolating of the conclusions. For example, India telecom boom should be read in the context of rural teledensity just being 32%(versus urban of 140% or so).
• Focus on outcomes not just on spending:- Especially in case of data on spending, avoid the tendency to mistake spending for outcomes. Try matching the spending to the outcome(say education and pass rates, health spend and mortality rate/life expectancy). For share price data, that may mean investor participation, retail folios etc

Not documented equals not done-leads to great quality or checkbox mentality?

While reading professional standards, it struck me that the basic premise of documenting a professional services engagement is that in case the professional is sued for malpractice, the working papers will prove the extent of, and diligence put in, the work done. Of course, it has positive uses also such as helping plan the assignment, building the knowledge bank for recurring assignments, and infusing a general discipline into the whole process. Also, there is always merit in having a checklist to avoid making mistakes during complex processes. Hence, I cannot reasonably argue the merits of following a checklist/process. The only issue is whether the person performing the assignment, will consider the checklist/standard as the end of his responsibility, given the protection from negligence lawsuits filed by disgruntled clients.

So now we are at a corundum. The protection from negligence lawsuits is the trigger to motivate professions to comply with service quality standards. But senior professionals themselves write the standards, so chances are they would incorporate recent memory, and rarely include future looking requirements identified from academic research. For example, India is going through several paradigm shifts in terms of introducing IFRS/XBRL/e-filing. But while ICAI has been quick to clamour for new professional work in these areas, it has been slower to issue fresh/modified guidance. And this is not a new phenomenon. New standards are often issued after the horse has bolted from the stable(viz scams/issues/controversies). For instance, while complex financial instruments valuation made the headlines in 2008, their formal auditing guidelines were published only in Oct-11. So for mistakes done before that, the auditor could claim protection. And some standards have been issued only when the ICAI has been dragged kicking and screaming. For example,the reporting standard on segment reporting/cash flow statement needed SEBI to step in, and cut the red tape.

So how do we motivate professionals to go beyond the minimal standards? There could be a best practices repository, and the professional submitting the most/top rated ones could be awarded. Or there could be awards for uncovering the highest amount of scam, mistakes, issues(subject to client confidentiality agreements) etc. Also, the data base could have a weightage while selecting auditors for key assignments and complex work. Lets see if this idea sees the light of day in my life!

Few Business systems see genuine goal subordination

Today, I opened a systems textbook and noticed the idealistic definition of system where a set of interrelated and interdependent sub systems, work together to achieve a common goal. And perchance if the sub system's individual goal(s) conflict with the overall goal, then the overall goal takes precedence. For instance, if the company's goal is to achieve maximum profitability, then it may conflict with an Operations Goal of 100% stock-in/Sales goal of 100% client coverage/Production goal of 100% capacity utilization. But it is naive to expect the respective functional heads to bend over and take it, just because of an idealistic sense of honour. After all, if their function underperforms, they can kiss their bonuses/promotions good bye, specially in the pay for performance mindset of today. Hence, controls and systems are set in place to ensure that sub units act as per the invisible hand to achieve overall goals. For instance

• Top Down Planning:- This ensures that sub units can only plan and execute within the overall boundaries laid down by management, consistent with superior goals
• Profit Centre and Transfer Pricing:-Profit Centre approach can ensure individual sub unit optimization, but may affect overall margins. That is why transfer pricing sets ground rules and dispute resolution for interactions between units, to ensure that the squabbles do not harm the organization too much.
• Cross Selling:- An acid test of the extent of integration between sub units, is the extent of cross selling done by SBUs. This metric is important especially for banks. Hence, encouraging cross selling will improve peer interaction, and boost overall profitability.
• 360 degree appraisal:- If peers have a role in appraisal, then atleast more civil behaviour and bother for overall goals will ensure.

But the best laid plans often fail. Top down planning without a strong internal audit function, often remains a paper tiger. Transfer pricing can be tweaked by smart management accountants, or rendered useless by faulty data/analysis. And cross selling is not always in the SBUs control, and may lead to misselling/fracturing the primary unit client relationships. These things are a reality of life in organizations, and how they are handled often determines 

What gets measured gets done-some unintended consequences from poorly designed metrics

Whether as students, employees, business persons or investors; we are measured by, and measure others by, metrics. The standard rationale for that is to use a standardized way to be more ‘objective’. But wrongly designed metrics have unintended consequences 

• ·         Research/Industry Interface/Foreign experience/Perception being given weightage in B-school rankings encourages colleges to encourage creation of poor quality research, invite industry personnel to ‘sell their company’ on campus under guise of industry interface and market the foreign student exchange as a fun trip instead of an educational experience.
• ·         The expectation to beat the street consensus EPS was a factor pressuring CEOs/CFOs to cook the books and focus on that metric, to the exclusion of cash flows, profits etc
• ·         Customer service has been given lip service in Indian service organizations(so ironic that the same outsourcing companies who win rave reviews from foreign clients do not seem giving a damm for their Indian clients), and now the regulators(IRDA/SEBI/RBI) have had to step in. If the investors/regulators/’Best Company Rankings’ had taken this factor into account earlier, surely there would have been some improvements.

So how to get it right? Like everything else in life, this is an art. But some get it right, by aligning the metrics to its strategy. For instance, the premier engineering conglomerate L&T has expressed its intention to divest some subsidiaries by 2015. Hence, the compensation plan of its top executives adjust the reported profits to exclude any divestment gains/losses, which are not wholly decided by them.

Takeaway:- List the Critical Success Factors, find the performance measures needed to address them, and then numerically express those measures as KPIs. And then you MAY begin to get it right.