Saturday, October 31, 2009

A primer on Analysis of Banks – Part 1

Recently I had a chat with one of my fellow junior..He asked me what is the significance of ratios like “D/E , Accounts payable , Accounts receivable ratio” for a bank…I had to elaborate what we learnt in Commercial bank management course in about half an hour…Here is the summary of our discussion

Function of a bank: Every organization has some reason for its existence. They benefit the owner, his relatives and to a certain extent a society at large..But the existence of banks is driven by the societal motive (Not that every owner wants to do it so, but the very design of a banking system makes it so!!)

If I want to put in simple terms the role of a bank is to borrow money from who have surplus and lend it to somebody who has deficit. This very definition of the bank makes its financial analysis different from companies in all other sector..

D/E ratio does not have any significance: How often have we heard analysts cribbing about the fact that the D/E of the company is almost reaching 1 which is a danger sign..But have we ever thought about the D/E of a bank..It’ll be approximately 10-12..Thats a very big number..It means the owner has put around 8 rs from his pocket and has borrowed 92 rs from outside..

As I have already said this is the business of a bank..They borrow it at a cheaper rate (through the deposits – fixed, floating …inter bank borrowings, borrowing from other agencies, borrowing from RBI) and lend it at a little higher rate..Hence they make a profit margin out of this transaction.

So can a banks always borrow and lend (with no money of its own) : Luckily there are stringent regulations (??!!) concerning the amount the owner of the bank needs to infuse in order to be complaint . We call it the capital adequacy ratio. Capital adequacy ratio, in simplified terms , is the amount of equity that a bank needs to hold given its balance sheet size..Now the Basel II regulation for banking assigns risk weights according to asset subclasses and according to the risk weighted assets you have to maintain a capital..and this ratio has to be atleast 8%...Let me clarify risk weight with an example..If the bank lends out to a AAA borrower like Infy it entails a risk weight of 50% meaning if he gives out loan of 100 rs to Infosys then the bank need to have 4 Rs(8/2) as equity capital..In the same way if it lends a housing loan , which entails 150% risk weight then it has to maintain 12 Rs(8*1.5) for every 100 Rs of loan it lends out..With every default that happens it is the equity capital which absorbs the losses..Hence if the equity capital goes below zero then the bank is said to be bankrupt(even if the bank might have 90% of its good loans still in the market)..

P/E ratio will also be closer to one: Turn into any newspaper and the first ratio that an analyst speaks about is the Price earnings ratio of a company..I was surprised yesterday to see a company with a P/E of 800..A company quotes at a P/E at such high levels because of the expectations of its future earnings..It means that the assets in the balance sheet of the company are worth much more than what is quoted..But if we observe for a bank it’ll be closer to 1..This is because of the fact that the bank has just interest income and interest expense..Hence there is nothing which is being quoted at a wrong value..All the assets and liabilities are quoted at proper values in the balance sheet(I don’t want to confuse you with the HTM , AFS , HFT concepts)..

Hope the part 1 has brought in some clarity..I will discuss some of the major risks faced by a bank and concepts like VAR in the part 2….

Sunday, October 11, 2009

Risky Assumptions - Lessons on Risk management learnt from the financial crisis

Through this article I highlight some of the assumptions on risk that investors, companies made which has terribly backfired on us/them during the recent financial crisis

Mean reverting characteristic

The first assumption in any standard book on risk management is the mean reverting characteristic of the risk variables..Let me explain this with the help of my favorite variable sigma (standard deviation)

When companies underwrite on securities(even during volatile times) they believe that the volatility(SD) will return to the long run average…Quite a decent assumption as long as the companies are able to stand the test of time..For example some of the companies/traders (who were short) were completely wiped out on a single day when the UPA government was elected because the volatility increased to a life time high of 75(VIX figure – source: NSE)

In the time of crisis the company’s liquidity takes a big beating which gives them no other choice but to wind off their positions at big losses.

Again I’m not completely supporting the mean reverting characteristic. During complete structural changes in the financial market in a country the variables change a lot and it will never come back to its historical mean.

Do we really STRESS the conditions?

The regulators and companies were smart enough to devise a system(even before the crisis) called as stress testing wherein they will simulate extreme scenarios and try to see what would be the impact on their balance sheet/profitability..

But what we/they failed to understand was the behavior of the market when conditions are stressed..So we merely do a scenario analysis of projecting different variable to extreme cases (example: Interest rate moving by 5%, volatility jumping by 100% etc)…Recently my friend gave me a hypothetical scenario – a really stressful situation and asked me how will I behave in that situation …My answer was ‘I don’t know’..He was surprised at the answer and asked me why I myself did not know..Simple because I have not experienced such a situation in my life and I myself can’t predict my behavior under such extreme conditions..So I said since I’m a very kind and nice person during normal circumstance does not mean I’ll be the same forever..Situations can drive a person crazy…Projection of the behavior is not the way out here..
Lot of the assumptions that we have taken during the normal market conditions fail terribly during stressed condition..One such important failure was the increasing asset correlations. Any global manager with significant exposure in various countries would have thought his positions to be safe because of the diversification..But during the financial crisis the asset correlations increased substantially that all their assumptions regarding correlation went for a toss..

United we stand – Says different risks

True to the proverbial statement which says ‘history always repeats itself’, there have been various individual cases of bank failure in the past due to individual risks which triggered the other risks and caused the collapse of the entire bank. For example it was a rogue trader (operational risk) who collapsed the billion dollar organization ‘Barings’ in no time. It was unacceptably higher levels of hedge fund investing (Liquidity risk) which triggered other risks and eventually caused the collapse of LTCM (long term capital management- A very famous hedge fund).It is entirely the banks and the financial institutions to blame who did not give the due respect for such historical incidents and went ahead doing whatever they were aggressively doing. The root cause of these problems lies in the fact that the organization failed to see the risk management from the enterprise perspective. Hence we infer that though there are different names for the risks (market risk, liquidity risk, credit risk, reputational risk) – they generally occur together

We use VAR:(Value At Risk)

VAR – Simple but an effective tool in identifying the pile of risk that you are sitting upon..Many companies were proud about the fact that they were using complicated models of VAR as the risk management tool and they knew exactly their risk..

But the problem with VAR is that it does not focus on extreme events..Regulators for banks say that the banks have adequate capital for their VAR with 99% confidence level…But what if the risk is only after the 99%..What if the magnitude of risk is exponential in nature after that 99%..Go ask any option writer and he’ll be able to explain this better (I luckily/unluckily being one among them)..

So using VAR is just a hygiene level..If companies want to identify the real risks they need to go much beyond what VAR suggests!!