Posts

/by

What-If Analysis In The Decision-Making Process – Test Your Hypothesis

,

Performing what-if analysis is an integral part of both the A/LM and budget processes. When used correctly, what-if analysis is a powerful way for decision-makers to understand the impact of items under consideration in real-time. The challenge is that often people dive right into modeling and results, producing a less than optimal process. Consider applying a scientific method to the what-if analysis to help strengthen the decision-making process.

The scientific method is essentially a hypothesis-driven methodology. Strong hypotheses lead to expectations either supported or refuted by analysis. What does this all mean? Well, it isn’t as intimidating as it might sound. From a financial modeling perspective, it means don’t just blindly rely on model results.

To help explain this concept further, consider a $1B credit union evaluating a strategy of moving $10M from overnights into 30-year fixed-rate mortgages:

What-If Analysis Test Your Hypothesis as Part of Decision-making Process

 

Before performing a what-if, the scientific method suggests that you first ask what you expect the results to look like, and then create a hypothesis. Start broadly with what you generally expect to happen to earnings in the current rate environment and the risk. In this case, the shift from overnights to mortgages should help earnings in today’s rate environment, adding risk as rates rise.

After identifying the broad expectation, take the next step and do some rough math to estimate the return on assets (ROA) impact of the what-if. Here, a $1B institution testing a strategy of moving 1.00% of its assets could expect a 3-basis-point improvement in the initial ROA (1% of assets multiplied by a 3% increase in yield):

What-If Analysis Testing Your Hypothesis as part of decision-making process

 

On the risk side, you can do the same with the impact to the net economic value (NEV) dollars since understanding the valuation impact is relatively straightforward. Overnights are at par in all rate environments while brand new 30-year mortgages devalue about 20% in a +300 basis points (bps) rate environment. Therefore, you’d expect to see a roughly $2M decrease in your NEV dollars in the +300 bps rate environment:

What-If Analysis as Part of the Decision-making Process Test Your Hypothesis

 

Analysis and observation are the next important steps in the scientific method. Run the what-if through the model and analyze the results in comparison to your expectation and rough math. Do the results of the what-if validate the hypothesis and, if not, why?

Periodically, results may not match up with the hypothesis, which is okay. It doesn’t necessarily mean the model or the hypothesis is incorrect. There could be other factors impacting the what-if. However, it is important to figure out why the results do not match up, especially if the difference is due to an input error.

For the example above, consider some of the following questions that could affect the what-if, causing the hypothesis and results not to match:

  • What was the credit risk assumption?
  • Will additional operating expenses and/or marketing dollars be needed to attract the growth?
  • Did we incorporate any fee income for the closing costs?
  • How long will it take to increase the portfolio $10M?

 

When it comes to the what-if process, shortcuts should not be taken. Always create an expectation internally before relying on model results. Depending exclusively on model results puts the user at risk of input errors and/or an inability to effectively explain what-if results.

/by

Examiner FAQs

We frequently hear about examiner inquiries regarding non-maturity deposit assumptions in credit union A/LM models.  The question is usually along the lines of, “what are the non-maturity deposit assumptions used in the A/LM modeling and how were they determined?”

Non-maturity deposit assumptions include pricing sensitivity and withdrawal sensitivity.  When it comes to pricing sensitivity the modeling needs to have assumptions about what the credit union thinks it will pay on its non-maturity deposits in different rate environments.  While there are many things to consider when it comes to pricing, one approach is to look at how the credit union has priced deposits in the recent past.  Short-term rates, which influence deposit pricing, were around 1% in 2003-2004 then rose steadily to about 5% in the 2006-2007 timeframe before dropping to the historically-low rate environment of the last few years.  A good starting point is to base your pricing assumptions on how you actually priced during this rate cycle.

Withdrawal sensitivity models the behavior of members who move their funds from lower-paying deposits (like regular shares) to higher-paying ones (like CDs) when presented with an opportunity to do so.  This behavior is much more difficult to observe than how deposits were priced in different rate environments.  In fact, in the IRR Questionnaire, NCUA says, “The uncertain timing of NMS account cash inflows and outflows can make treatment challenging.  It is not possible to predict with certainty what future balances in non-maturity accounts will be, how long they will remain open, or what future rates will be paid to members on these accounts.  Even when CUs study member behavior, or contract with vendors to perform such a study, substantial uncertainty remains.”

Similar to the approach with pricing, it is possible, however, to observe how your credit union’s deposit balances responded during the last rate cycle.  Reviewing balance changes, especially during the period of 2004-2006 (when short-term rates were increasing), can provide a basis for withdrawal sensitivity assumptions.  Still, there is nothing in recent history that replicates this extremely low-rate environment and there are valid concerns that member behavior in the future may be very different than in the past. Movement of funds to higher-paying deposits, or having to replace funds that are leaving with higher-paying deposits, can dramatically increase a credit union’s cost of funds.  As with the pricing assumptions, it is a good practice to stress test these assumptions by asking, “what if our member withdrawal is X% (for example 50%) greater than assumed?”

One last consideration:  the type of analysis you are doing. Deposit pricing assumptions are needed for static and dynamic balance sheet income simulations, NEV and long-term risk to earnings and net worth simulations.  Withdrawal assumptions are needed for NEV and long-term risk to earnings and net worth simulations.  While dynamic simulations wouldn’t necessarily employ withdrawal assumptions, it is possible to model changes in the deposit mix. Static balance sheet income simulations, by definition, ignore this threat by assuming that deposits never leave and that members never act in their best interest by moving to higher-paying deposits.

/by

How Do You Know Your Modeling Assumptions Are Right?

,

Interest rate risk modeling requires the user to make assumptions about member and management/board behavior.  For example, some members will pay back their loans ahead of schedule and the rate of prepayment will increase if rates fall and decrease if rates go up, especially on mortgage-related products.  Likewise, the incentive for members to move some of their deposits from lower-cost products, such as regular shares, to higher-yielding CDs increases when rates go up (note: static income simulation ignores this important aspect of IRR management).  Assumptions must be made about the likelihood of these events occurring in the future.

Has your examiner asked “how do you know your modeling assumptions are right?”  If so, you answered correctly if you said you don’t.  Assumptions, as the name implies, are assumptions.  They are not facts.  A prepayment speed on a loan can be calculated with certainty when the prepayment occurs.  To put that prepayment speed into a model assumes that behavior will continue, but it may not.  How that behavior changes when interest rate changes is yet another assumption.  Despite this, the assumptions are an important and necessary part of modeling a credit union’s IRR exposure.

The new IRR rule, 12 CFR Part 741, Interest Rate Risk Policy and Program, says assumptions should be “reasonable and supportable” and that credit unions should “assess the sensitivity of results relative to each key assumption.”  We agree with these comments and work with our clients to develop reasonable assumptions, yet we realize there is no right assumption.  Therefore, we conduct literally thousands of stress tests annually to help our clients understand the impact of their assumptions to their IRR exposure.  For example, we often stress test early withdrawal speeds by making them 50% faster than what is in the base simulation.  Whether you are a client of ours or not, a recommended approach is to stress test assumptions to see how they may impact results.  Assumptions that have a material impact may require additional research and refinement whereas others may not.

The bottom line is stress testing and documenting—followed by assumption adjustments if appropriate—should help to satisfy the “reasonable and supportable” component of the new rule.

/by

Hot Money In Waiting

, ,

Many credit unions continue to see higher levels of deposit growth despite lowering rates—and even a rising stock market—as members choose safety and certainty over return.  While the flight to safety has been discussed at length, the gradual shift from CDs to share products has received less attention.

Over the past 6 to 12 months, many credit unions have seen decreases in CD balances with a corresponding increase in share balances, in addition to the increase in overall deposit balances.  Such a trend suggests that members are willing to park their money in a lower-paying share account rather than lock their money up in a higher-paying CD in order to have the flexibility to reinvest their money when a better alternative presents itself.  As a result, these balances could be hot money in waiting.

Credit unions should evaluate the change in their liability mix over the last year and consider how any shifts might affect their liquidity concerns, product needs and cost of funds going forward—especially in different rate environments.  Likewise, it would be prudent for institutions to incorporate this additional rate sensitivity into their modeling, particularly if a clear shift can be identified.  Modeling the hot money as 30-50% more sensitive than other share balances is a good place to start.

Whether the impact is large or small, the credit union will be better positioned to handle the reinvestment of hot money should it occur.

/by

Planning For PLL

,

Given the economic experience of the past couple of years, many credit unions were forced to beef up their loan loss reserves.  Now, as things appear to be getting better, we are seeing that some credit unions feel they are “overfunded” and are not adding to their reserves.  While this may provide some temporary relief to earnings, credit unions cannot plan their long-term business models on the fact that there is no PLL expense or, in some cases, that the expense is negative.

Credit unions have worked hard in this environment to define their target markets, focus their efforts toward them and have learned to do things more efficiently. They should enjoy this brief reprieve.  However, credit unions should not become complacent.  Rather, they need to continue efforts to position themselves to be better and stronger in the future.

As far as modeling goes, assume that the PLL expense is at the level expected after the allowance for loan loss reaches an adequately funded level.  This will provide a more realistic picture of long-term earnings.  From a risk-management perspective, consider the experiences from the last couple years in making assumptions about worst-case credit risk exposure, not only from loans but also from investments.