Manage your bank, not your model

Can you clarify what no-growth means < from the 2012 FAQ

“You can model your risk to be anything you want it to be.”  That’s a pretty popular saying among modeling skeptics.  They recognize with “forward-looking” stress-tests there is a real danger of manipulating the results based on what you think will happen in the future.  Loan and deposit growth, new product pricing, or changes in balance sheet mix may all be changes you expect to happen in the future but…what if they don’t?  Given that we’re generally pretty bad at predicting the future it’s prudent to run a stress-test that assumes your balance sheet mix and volume will remain pretty much the same.

Question from the FAQ:
8) When no growth scenarios for measuring earnings simulations are mentioned, can you clarify what no growth means?

Answer from the FAQ: 
“No growth” refers to maintaining a stable balance sheet (both size and mix) throughout the modeling horizon. Financial regulators are concerned that including asset growth in model inputs can reduce the amount of IRR identified in model outputs. For example, if model inputs predict significant loan growth occurring after a rate shock, new loans are often assumed to be made at higher interest rates. This has the effect of reducing the level of IRR identified by the model. If this assumed growth does not occur, the model would underreport actual IRR exposure.

Institutions should recognize and understand how growth affects model output. Management should run scenarios that maintain the balance sheet constant across the simulation horizon. These types of scenarios help highlight the current level of risk in the institution’s positions without the effects of growth assumptions. As a sound practice, management could contrast the “no growth” scenario with scenarios that include growth assumptions to highlight how future growth may change the institution’s risk profile.

This no-growth scenario is sometimes also referred to as a “static” forecast.  This is in contrast to a forecast that might include some sort of growth which is often referred to as a “dynamic” forecast.  The original IRR advisory addressed the difference between the two.  Back in September I offered further clarification:

Most people…assume that static means “no-growth” and dynamic means “growth”…That’s almost right. It is generally accepted that static means “no-growth” or a constant balance sheet. But dynamic doesn’t necessarily mean “growth”. It really means “changes” in the balance sheet.
What’s the difference between a static versus a dynamic forecast? | Sept 2011

Should all institutions analyze risk other than repricing risk < from the 2012 FAQ

YES! I’m really surprised the regulators didn’t just simply leave the answer at that.  Now that I look at it - they didn’t even answer “yes”, which is sad.  Any bank that isn’t already identifying option risk and basis risk in their IRR models is really missing the boat.  In the JPS from 1996 the agencies did a pretty good job of outlining the five types of risk they wanted bankers to identify – repricing/maturity risk, option risk, basis risk, yield curve risk, and price risk.

I think this is one of the most poorly worded questions and answers in the FAQ.  I’ll tell you why after we revisit the actual text from the document:

Question from the FAQ:
6) Should all institutions analyze risk other than repricing risk (i.e., non-parallel yield curves, basis risk, and options risk)? If so, how often should risk analyses be run?

Answer from the FAQ:
The advisory states that the types of stress scenarios depend on the risk profile of the institution and the complexity of its structure and activities. All institutions are expected to run these types of scenarios periodically to fully identify significant positions in the four components of IRR: repricing mismatch, basis risk, yield curve risk, and options risk. Institutions should conduct analyses for basis, yield curve, and options risk as necessary, depending on the complexity of activities and risk profile. Generally, these analyses should be run at least annually, or when the risk profile of the institution has changed (for example, because of acquisitions, significant new products, or new hedging programs). Ideally, these analyses would be conducted for earnings calculations as well as economic value of capital measurements.

If an institution’s risk profile shows a significant sensitivity to one of these risks, this scenario should be included in the regular monthly or quarterly IRR monitoring. For example, if an institution maintains a relatively short net duration balance sheet, but uses two indices to price assets and liabilities, a basis-shift scenario may identify IRR exposures that otherwise would not be detected in an interest-rate-only scenario. For institutions that price assets primarily from long-term rates, and liabilities from short-term rates, a change in the shape of the yield curve typically would be a more appropriate scenario.

You can get a better appreciation for this question if you understand that most banks only run instantaneous parallel shifts when measuring IRR.  A better way to ask this question would be:

Should the bank run additional kinds of stress-tests other than the instantaneous parallel shift (i.e. non-parallel yield curve shifts or gradual rate ramps)?  If so, how often should they be run?

The answer in the FAQ is reasonable – run them at least annually or when the risk profile of the institution has changed.

What is confusing about the question and answer as they are written is they imply that banks which only run parallel shifts are only identifying repricing risk – which is just plain wrong.  Nearly every A/L model can change the nature of cash flows when appropriate i.e. slow-down or speed-up prepayments, call a bond or not (also step-ups), debt conversion, and early withdrawals.  These are all examples of option risk and it is very common to see these accounted for in a parallel shift analysis.

Typically the standard parallel shift also identifies a major source of basis risk.  Banks will typically model a beta or lag in core deposit pricing.  This doesn’t fit the strict definition of “parallel shift” but it is a very common behavior that most banks build into their models.

So most banks already identify repricing risk, option risk, and basis risk in their models.  I think what the examiners are driving at is that non-parallel shifts and rate ramps can shed a different light on things.  You may see behaviors that you might otherwise not see in the parallel shift.  Running these non-standard kinds of interest rate stresses may be useful on an infrequent basis.  Though most community banks don’t usually discover anything new about their IRR (that they didn’t already know.)

Grasping the fundamentals

There’s been a recent effort by regulators to focus industry attention on interest rate risk.  They have given quite a bit of attention to these three considerations:

1. The interest rate scenario used to measure the effect of changing rates;
2. The asymmetrical rate of sensitivity that results form certain bank products when both rising and falling interest rate scenarios are considered (a.k.a. option and basis risk); and
3. The treatment of non-maturity deposits (demand, NOW, savings, and MMDAs.)

I’m certain there would be debate about which is the most important.  But it’s hard to deny, when dealing with these considerations, the method you choose can significantly impact your stress-test results.  The regulators are well aware of this and they have been for quite some time. 

Here’s a trivia question for you – when was the first time these three issues were spelled out in an official communication from regulators?

a) 2012 Interest Rate Risk Management: Frequently Asked Questions (Joint Policy Statement)
b) 2011 Supervisory Guidance on Model Risk Management (Fed & OCC)
c) 2010 Interest Rate Risk Advisory (Joint Policy Statement)
d) 2000 Guidelines for Model Validation (OCC)
e) 1996 JPS on Interest Rate Risk
f) some other document

If you guessed “e) 1996 JPS on Interest Rate Risk,” you would have guessed the correct decade, but still the wrong answer.  The correct answer here is “f) some other document.”

The regulators actually listed these three issues as, “Major Considerations in Measuring Interest Rate Risk,” in a notice of proposed rule making back in 1993 - that was nearly nineteen years ago.  (Actually this is just the earliest appearance I could find.  They likely appeared prior to this sometime in the 1980’s – I just haven’t found it yet.)  Here's a link to the entire document.  It’s fascinating to reread this document after all these years. It’s pretty clear many of the questions about measuring IRR that we’re struggling with today aren’t really new questions at all. Many of them were raised and discussed years ago.  The debate continues about these three elements…even nineteen years later:

1. What is the appropriate amount to shock rates?  Maybe we should also rate ramps and/or twist the curve?
2. What is a reasonable base case prepayment speed to use?  How should that speed change when rates change?
3. Should non-maturity deposits be modeled long-term or short-term?  Is there a process to find out?

I think it’s safe to say that these are indeed fundamental assumptions for measuring IRR.  Your IRR measurements are significantly influenced by how you handle these assumptions. Running a larger number of tests (e.g. shocking 500bp or 600bp) won’t really matter if you’re not certain that shocking is the “right” test in the first place.  Forecasting over longer or multiple time frames (e.g. 12 months, 24 month, or longer) doesn’t really help if your assumptions about non-maturity deposits are suspect.  The reality is we still don’t fully understand what the “best” assumption is for any of these (especially the non-maturity deposits assumption.)  We have some reasonable approaches to modeling them, but none can be considered the “right” way. How they are handled is every bit as “Major” today as it was nineteen years ago.  Running more and more stress-tests will not diminish their importance (or the fact that they are still assumptions.)

This is something the industry (regulators, bankers, and analysts) should keep-in-mind before they try and fit stress-testing results into nice neat buckets called “base”, “adverse”, and “more-adverse”.  They should also be wary of making stress-test results publically available, especially when they are based on such non-standard or “fuzzy” assumptions.  These are indeed “major” considerations, but we’re not even close to definitively dealing with them yet.

Summarize or benchmark results

How much risk is too much?  That’s the most common question I get from our clients when they’re reviewing stress-test results.  I addressed this question somewhat back when I wrote about setting appropriate ALCO policies.  It’s difficult to answer the question generically and out-of-context.  The common answer is usually “it depends…”.  Every bank has a different risk tolerance.  Every bank operates in a unique market.  Every bank has a different capital position. etc.  Identifying what risk we’re talking about and using familiar systems to measure it are good first steps to take towards answering the question “how much risk is too much?” But once we’ve taken the measurement it’s still often difficult to decipher.

For example, what if I told you that your bank’s net interest earnings at risk was –2.81%?  Could you tell me if that is a high or low level of risk?  If you are the bank CEO or CFO perhaps you could, but what about the least financially savvy member of your board?  Could he/she tell?  Over the years I’ve observed that most community bank board members are uncomfortable with this question.  Even senior executives, auditors, and examiners often have trouble qualifying risk measurements like this.  Being unable to determine how much is too much is a problem because every piece of regulatory literature to date says that “…risk limits should be approved by an institution's Board of Directors…”.  How can they set effective limits if they can’t decide how much is too much?

One of the easiest ways to address this problem is to summarize or benchmark the stress-test measurements.  We need to recognize that not everybody who reviews the measurements will be an expert.  Most folks will need some sort of guideline or context which will help them interpret the measurement.  Without a context the measurement is just data.

Best selling author and marketing guru Seth Godin has written a number of great articles about the need to establish a context.  A recent one entitled, “The power of visualization", has this to say about data and information:

Data is not useful until it becomes information, and that's because data is hard for human beings to digest…We repeatedly underestimate how important a story is to help us make sense of the world.
Seth Godin’s Blog | October 19, 2011

In our A/L BENCHMARKS product we’re constantly looking for ways to turn data into information.  Here’s how we present the –2.81% earnings-at-risk measurement in our Board Report. This particular graph/narrative helps communicate several things:

1) The current position: the current level of risk (as of 2nd-2010) is –2.81% and it occurs in the rates down (“DN”) environment.
2) The past position: the level of risk has fluctuated some over the past six quarters, but in-general it looks pretty consistent i.e. not much volatility.
3) The average: on average, bank’s this size have a higher level of risk, closer to –6.75%.  The bank has been consistently lower than average.
4) The variability:  the average is obviously comprised of many measurements, some higher and some lower.  What is the variability around the average, or what is the typical range of measurements?  That’s shown by the box that looks like it’s floating behind the bars.  The bank’s –2.81% is even lower than what is typical.
5) The story: not everyone digests numbers and charts.  Some folks prefer a narrative.  On top of the chart is a text explanation of the data.

This presentation isn’t perfect.  I’m sure there are business communication experts that could point out several flaws that, if corrected, would more completely and precisely represent the data.  But, that’s not my purpose – I want to summarize and benchmark the data.  I’m trying to give experts and non-experts alike a context.  I’m trying to turn stress-testing data into meaningful information.

(This post is part of a series entitled “Stress-testing Do’s and Don’ts”)

The +400bp shock to come – maybe

No discussion about measuring interest rate risk is ever complete without a debate about the “most appropriate” level of stress to use.  While the +100bp, +200bp, and +300bp parallel shifts remain pretty much the standard, the +400bp shock is gaining popularity.

There are certainly historical examples of large (almost instantaneous) swings in rates.  Look back at November 1980 when the average Fed Funds rate was 15.85%.  By the end of the following month the average was 18.90%, basically an increase of +300bp (in one month!)  Then there was the precipitous drop starting in November 2000 when average Fed Funds was 6.50%.  By the end of the following year the average rate had dropped over –450bp to 1.82%. 

Now that we’re currently sitting in one of the lowest rate environments in history there is wide-spread speculation that rates, when they do start to move, will rocket upward.

Even several members of the Fed predict this kind of dramatic change.  In the Projection Materials from the January 24-25, 2012 meeting at least six board members are predicting a rapid rise in rates sometime in 2015 or 2016.  You can’t track the movement on the graph of each member specifically, but you can certainly see the range.

Seventeen policymakers’ predications were recorded (five board governors and twelve regional Fed Bank presidents).  Six of them project anywhere from a +350bp to +425bp in 2015/16.  Six out of seventeen, that’s more than a third of the predictions provided…certainly something that can’t be ignored.

Of course all these predictions come with the standard caveats and conditions:

Appropriate monetary policy, by definition, is the future path of policy that each participant deems most likely to foster outcomes for economic activity and inflation that best satisfy his or her interpretation of the Federal Reserve’s dual objectives of maximum employment and stable prices…These projections of the timing of the initial increase of the target federal funds rate and the path of the target federal funds rate are the ones that policymakers view as compatible with their individual economic projections.
Economic Projections of Federal Reserve Board Members and Federal Reserve Bank Presidents, January 2012

Clairvoyants – the Fed reveals their rate forecast

clairvoyant (noun, plural clairvoyants)

• A person able to see things that cannot be perceived by the normal senses.
• A person able to foresee the future.

via Wiktionary, the free dictionary

In late 2011 the Fed decided to start providing more specific information about its interest rate expectations.  Well it’s here – in the Projection Materials from the January 24-25, 2012 meeting.

There’s good coverage of this on the web (there are already thousands of results when you Google it).  Today’s Wall Street Journal covers it pretty well:

Federal Reserve officials said they expect to keep short-term interest rates near zero for almost three more years and signaled they could restart a controversial bond-buying program in yet another campaign to rev up the disappointing economic recovery.
Fed Sees Low Rates to 2014 | Wall Street Journal | January 26, 2012

The WSJ graph (shown at the bottom of this page) is easier to understand than the official one published by the Fed. 

I worry about this projection for two reasons:

1. There is a real possibility that the Fed will damage their credibility.  The minutes released from Fed meetings back in 2006 illustrate pretty clearly that they failed to really see what was coming.
2. They project an illusion of control over something that two, three, even four or five years from now they really have no control over at all.  Most people in the financial sector know better, but that doesn’t really matter.  Even the WSJ article uses phrases like (emphasis mine), “…businesses see the Fed's assurances”.  The words “assure” and “assurances” are thrown around quite a bit.  There are just too many variables.  If they had that much control – why couldn’t we avoid the current mess?

In the press conference Bernanke provides his requisite CYA, “Our ability to forecast three and four years out is obviously very limited.  It's certainly possible that we will be either too optimistic or too pessimistic.”  These are the kinds of statements that no one ever hears – instead they just judge your forecast as “right” or “wrong”.

Modeling new strategies < from the 2012 FAQ

I have to admit, when I first read question number two I was a bit skeptical that this was even a “real” FAQ.  It seemed more like an attempt to inject some formal guidance into the document disguised as a real question from bankers.  If your model’s not capable of measuring the new risks you are taking then – yes – that will raise concerns. Duh.  Here’s the full text from the FAQ:

Question from the FAQ:
2) If an institution implements a new strategy and later finds that its IRR measurement model cannot capture the risk exposure, could this raise significant supervisory concerns?

Answer from the FAQ:
Yes. All potential risk exposures, including IRR, posed by new products or strategies should be considered as part of the due diligence for any new strategy. If a new strategy involves IRR that cannot be adequately captured by existing measurement processes, steps should be taken to ensure this risk can be adequately measured before implementation of the strategy. The cost of measuring the change in exposure from a new product or strategy also should be considered an essential part of the due diligence process.

For example, if an institution were to implement a leverage strategy using highly structured liabilities to fund fixed-rate mortgage investments or whole loans, this type of strategy could introduce a significant level of option risk to the institution’s IRR risk profile. If existing IRR measurement tools do not adequately capture the potential volatility in cash flows and rate adjustments from the newly acquired assets and liabilities, the model would not be able to adequately capture this option risk. Therefore, management would not be able to measure the IRR exposure accurately. This would likely be considered a management weakness, and corrective actions could include making the appropriate changes or enhancements to the model. In some cases where on- or off-balance-sheet items cannot be effectively measured in the primary IRR model, it may be appropriate to use alternative means to measure the risk in such products, where the alternative output is then incorporated into the primary model results (i.e., add-ons). Financial regulators expect risk managers to consider the ability of current systems to model risks posed by a new strategy in advance to understand how new products or strategies affect overall IRR exposure.**(see footnote)

**[Footnote] - Typically, institutions that have in-house or turnkey vendor models can generate alternate or “what-if” measurement scenarios outside of normal IRR exposure reporting. Institutions that do not have access to run their own models (those that completely outsource the measurement process) can use other means to estimate the risk of new strategies based on the size and complexity of the institution’s activities.

The main part of the answer essentially says nothing more than, “the complexity of your model should match the complexity of your institution”.  I think the footnote is the most important part here.  A better question might be, “Is it important for my model to have what-if capabilities?”  Answer: Yes.  However, I don’t like the wording of the footnote.  It suggests that out-sourced models don’t allow banks to run what-if scenarios.  For many out-sourced solutions that may be the case.  However, A/L BENCHMARKS clients certainly can run what-if scenarios.  If you are a client and you are considering a new pricing, investment, or funding strategy that will impact a significant portion of your balance sheet – please let us know.  We can show you the impact such decisions might have on your Earnings and EVE at risk.

(It also suggests that banks that own an in-house model are all capable of running what-ifs…their software may be, but often the in-house expertise is missing – see “The Worst of Both Worlds”).

Leverage existing systems

Given the amount of media coverage and popular interest in the bank stress-testing that was part of SCAP you might think that stress-testing was a relatively new idea to the financial industry.  A look at recent regulatory literature might even confirm that notion.  Here is a list of several documents that specifically mention stress-testing:

• 2006 Commercial Real Estate Lending Joint Guidance (FIL-104-2006)
• 2008 Liquidity Risk Management (FIL-84-2008)
• 2009 Basel Committee’s Principles for sound stress-testing and practices and supervision (BCBS155)
• 2010 Interest Rate Risk Advisory (PR-2-2010) (my thoughts and comments here)

However, stress-testing is not a “new” idea in the banking world.  Indeed the regulators introduced the notion of stress-testing more than fifteen years ago in their 1996 JPS on Interest Rate Risk.  While the document itself talked specifically about interest rate risk, it outlined several concepts related to stress-testing in general.  One interesting thing about the document is that the explicit phrase “stress-testing” didn’t appear anywhere in the document.  They made references to “stress scenarios” and “extreme conditions” but never called it stress-testing by name.

Assumptions about…new business activity should be reasonable and consistent with each rate scenario that is evaluated. In particular, as part of its measurement process, bank management should consider how the maturity, repricing and cash flows of instruments with embedded options may change under various scenarios.
1996 JPS on Interest Rate Risk

Make no mistake though the 1996 JPS was a great introduction to stress-testing.  In fact in the 2010 IRR advisory the regulators specifically refer the reader back to points made in the 1996 document.  Industry developments in Asset/Liability modeling after 1996 all started to focus on creating forward looking scenarios and projections to measure and monitor IRR per the suggestions made in this JPS.

After fifteen years it’s a safe bet that almost all banks are using some form of scenario stress-testing to measure and monitor interest rate risk.  In fact, IRR stress-test reports, or “shock reports” as they are commonly called, often come prepackaged with core banking systems and applications.  It’s a rare banker indeed who hasn’t seen an interest rate risk shock report.  It’s common place.

That’s why after hearing from several vendors that were part of the Stress-Testing for Community Banks forum and seeing the deluge of stress-testing marketing information in the industry I’m left scratching my head.  If you believe this new wave of vendors, stress-testing and the tools you need to do it are all “new!…new!…new!”  Quite often their sales pitches boil-down to this:

1. Looking at current/historical data is no longer good enough.  You need a tool that will project your future position to give you a more dynamic picture of your bank.
2. These projections need to be tested.  You need a tool that will apply certain stresses as variables to measure the amount and severity of your exposure.

These two things sound great, exactly the sort of thing banks need to do to follow good stress-testing practices.  But aren’t most banks already doing this with the tool they have in place to create a projection (e.g. a budget or strategic plan)?  Aren’t most banks already “shocking” these projections to measure the impact of changing interest rates (i.e. a stress variable)?  It seems to me that the modeling process most bankers have in place to measure IRR already does these things quite well.  Take measuring earnings at risk for example (see insert).  Why reinvent the wheel?

There is no doubt that the process would need to be tweaked somewhat to measure different types of risk (i.e. liquidity or credit).  But there shouldn’t be any need to purchase yet another tool to gather, model, and report data.  Banks already have tools like this in place and they should leverage these existing systems.

(This post is part of a series entitled “Stress-testing Do’s and Don’ts”)

Which IRR vendor models are appropriate < from the 2012 FAQ

For years bankers have been asking (pleading with?) their regulators to just tell them what model to use.  And for years regulators have answered something to the effect of, “the type and size of the model you need should match the size and complexity of your bank.”  Somewhat vague I agree, but I understand where the regulators are coming from.  They don’t want to let the banker off the hook, they don’t want to suggest a tool that doesn’t appropriately capture the bank’s risk exposure, and they certainly don’t want to be in the business of “suggesting” one vendor over another.  In other words, it’s up to the bank to determine what’s appropriate or not.

Interestingly enough the now non-existent OTS used to take the very opposite position on this by running their NPV model for every thrift.  This has now created somewhat of a problem for the OCC since they are now the primary regulator for thrifts.  Many thrifts have never run an IRR model themselves (in-house or out-sourced), and many lack the expertise to do so.

Question from the FAQ:
1) How should financial institutions determine which IRR vendor models are appropriate?

Answer from the FAQ:
Models can vary significantly depending on complexity, data management, and cost. Achieving the proper balance among risk positions, risk measurement processes, and cost is critical to a successful model risk management program. When creating an IRR model or evaluating third-party models, institution management should thoroughly assess the model’s ability to reasonably capture risks in the institution. Additionally, management should reevaluate the model’s appropriateness as risk positions, strategies, and activities change. When reviewing modeling options, management should at a minimum consider the following:

• The ability to reasonably model the institution’s current and planned on- and off-balance-sheet product types (on both income and capital valuation bases). Material positions in highly structured instruments or institution-specific products should be key considerations. The model should support the level of data aggregation and stratification** necessary to properly measure these types of products.
• The extent to which the model uses automated processes compared with manual procedures. Management should consider whether the model has automated interfaces with institution source systems. Management should also consider the cost, hardware and software requirements, staff resources, and expertise needed to run the model and integrate any separate (manual) add-ons (also see question #2).
• The level of model transparency and the adequacy and comprehensiveness of vendor model validations and internal control reviews (also see question #9).
• The level of vendor implementation and ongoing support received, including available training from the vendor.

To better control third-party model risk, financial regulators expect financial institutions to have sufficient in-house knowledge in case vendors or financial institutions terminate contracts for any reason, or if vendors are no longer in business. Financial institutions should maintain contingency plans for addressing how management should respond to such lapses in vendor support.

The bottom line is that the bank should be well aware of the capabilities and drawbacks of using a certain model.  If the model is too simplistic it is probably going to be missing some risks (the most commonly overlooked risk is option risk.)  But, you don’t want to over buy either, that is implement a model that is too sophisticated, or requires too much time and other resources to update.  There are just as many good and bad in-house models as there are good and bad out-sourced models.  Beware that you don’t fall into the trap of “the worst of both worlds.”

Believe it or not many banker’s don’t know if or how their model handles:

• Loan or MBS prepayments
• Bond call or step-up options
• Core deposit pricing (betas)
• Core deposit decay
• Deposit early withdrawals
• Debt conversion options

Just to name a few.  You should recognize that there are many poorly implemented and updated models (both in-house and out-sourced) that don’t adequately address these behaviors.  Just using a “reputable” vendor’s model is no guarantee.

---

** “data aggregation and stratification”…couldn’t they just have said “collect and organize data”?  It’s wording like this throughout the FAQ that makes it a bit hard to read.

Regulator’s FAQ on Interest Rate Risk

Last week the FFIEC released this FAQ for the 2010 Interagency Advisory on IRR. It’s a decent document that covers some important questions that I know clients and potential clients have had. However, it’s by no means a comprehensive document. I’ve developed a series of posts that comments on each of the questions for the benefit of our current and future A/L BENCHMARKS clients. Have they made things clearer? Can you expect anything different during your next exam based on this document? Let’s look at each question in detail (one per week starting 01/19/2012):

Risk Management/Oversight
1) How should financial institutions determine which IRR vendor models are appropriate?
2) If an institution implements a new strategy and later finds that its IRR measurement model cannot capture the risk exposure, could this raise significant supervisory concerns?

Measurement and Monitoring of IRR
3) What types of IRR measurement methodologies are institutions expected to use?
4) Should institutions with non-complex balance sheets use earnings simulations to measure risk to earnings?

Stress-testing
5) Should institutions perform rate shocks greater than +/-300 basis points?
6) Should all institutions analyze risk other than repricing risk (i.e. non-parallel yield curves, basis risk, and options risk)? If so, how often should risk analysis be run?
7) Should institutions establish board-approved thresholds for monitoring each stress scenario they run?
8) When “no growth” scenarios for measuring earnings simulations are mentioned, can you clarify what “no-growth” means?

Internal controls and Validation
9) Most institutions use third party tools to measure IRR. Can independent certifications/validations commissioned by model vendors satisfy supervisory expectations for model validations?
10) Can you provide some examples of effective back-testing practices?

Assumptions
11) Can an institution use industry estimates for non-maturity-deposit (NMD) decay rates?
12) Regarding deposit decay-rate assumptions, what are some examples of a “market environment in which customer behaviors may not reflect long-term economic fundamentals?”