Month: December 2021

(Part 3 of 3 on the NCUSIF , A Better Way for credit union share insurance)

In 2010 GAO reviewed the FDIC’s financial statements for 2008 and 2009 (report 10-705). The summary, What GAO Found, contained the following comments:

“Because of a material weakness in internal control related to its process for estimating losses on loss-sharing agreements, in GAO’s opinion, FDIC did not have effective internal control over financial reporting  . . As of December 31 2009 the DIF had a negative fund balance of $20.9 billion and it had a negative 0.39 percent ratio of reserves to insured deposits.”

This was the third negative position for the fund since deregulation. In the most apocalyptic and highly erroneous NCUA projections about the corporate crisis during (and after) the Great Recession, none projected the NCUSIF would ever be in a deficit position.

After these two significant  economic downturns in one decade, the NCUSIF stands tall, stable and resilient.  However current NCUA Chairman Harper has openly called for the NCUSIF to be changed so it can mirror the FDIC-an entirely different financial design.

His suggestions show a  lack of knowledge about how the NCUSIF has succeeded in an environment that has seen every  other federal deposit insurance premium-based system fail.

NCUSIF’s Record in Recent Crisis

The 13-year period from 2008 through 2020 includes two serious economic downturns.

The cumulative insured loss rate for credit unions during this period is 1.5 basis points(bps) as shown below.  The annual losses range from 0 in 2020 to 6.96 bps when the taxi medallion losses were recorded in 2017.

The NCUSIF’s financial design rests upon the 1% deposit underwriting which grows proportionately with the insured risk plus an equity cushion that has had a traditional cap of .30 basis points of  insured shares.

The board is required to manage this equity level between .2 and .3.  If the NOL is under 1.2%, then the board must present a plan to return to this range.  Above the 1.3% cap (or other limit set by the board to a maximum of 1.5%), the board must pay a dividend to credit unions to reduce the equity to 1.3%.

Loss Provision Estimates Way off Mark

The fund has never incurred losses close to this 10 basis point range.   However NCUA’s loss estimates have often been spectacularly in error.   This graph shows that net cash losses reported in audits have no relation to provision expense.

The next chart shows how the much the overfunding of reserves  compares to actual losses.

The yearend loss reserve has fluctuated wildly  exceeding subsequent actual cash losses by over 10 times in multiple years (1,000%).

These wide disparities between actual losses and provision estimates require subsequent reserve adjustments.  These create very misleading financial bottom line results.  The provision expense, subtracted from or added back to earnings does not present  actual loss experience in a timely and consistent manner.

The fund’s structure is more than adequate for insured credit union losses.   What is most ominous is that NCUA seems to have no consistent, objective and verifiable method for setting the loss reserve.  That is a management problem, not a fund design issue.

When the  annual operating expenses of 1.8 basis points are added to the 1.5 basis points, total costs average 3.3 basis points.  As outlined earlier, the most persistent increasing expense is NCUA’s overhead transfer, not the costs from insured loses. This expense should be more controllable but has in fact become an increasing funding source covering almost two thirds of NCUA’s annual budget.

What About Revenue?

Costs are covered by revenue.  In addition to the two costs above,  income is required to maintain the retained earnings portion of the NOL growing at the same pace as share growth to keep a ratio in the .20-.30 equity range.

The math is easy.   An example. Assuming $1.5 trillion of insured shares, a cash loss rate of 1.5 basis points requires $225 million of income.   As investments equal 1.3% of insured savings, the an investments yield 1.2% covers losses.  To pay the 1.8 bps overhead expense adds another 1.4% for a 2.6% yield to cover all costs in the most 13 years.

If  the long term growth rate is 6.5% of insured savings, an additional yield of 1.3% is necessary to maintain equity. Using just the last decades results would seem to require a 3.9% yield to sustain the funds NOL.

But this return on investments is not required in most normal years.  And a far lower yield, as in the 2021 of 1.23%, can still result in a very strong bottom line.  Here’s why.

First, average insured losses at 1.5 bps are much higher than the long-term historical outcome.  The 13-year experience reflects two unusual events: the Great Recession of 2008-2009 and the taxi medallion losses.   If those three years removed from the average, the average insured losses are far below 1 basis point.

In 2021, insured share losses have been zero and there have been net recoveries from previous loss expense provisions.  In 2020, there were nil insured losses. In the previous three decades prior to 2009, insured losses were often zero and far under the most recent period’s 1.5 bps.

Overhead expense can be controlled.  It has grown at the same rate as insured shares (6.5%) because NCUA has transferred a disproportionate amount of its budget increase in every year to the NCUSIF—over 62%.  If the NCUSIF expenses grew at the same percentage as NCUA’s overall budget, this would reduce the 1.8 basis point portion significantly.  Moreover the growth in the investment portfolio shows that in 2021, the yield needed to cover this year’s projected $195 million OTR is only 1 bps.

Finally the equity ratio is a range of 1.2-1.3.   It can and has historically varied between a low of 1.24 to the cap set by the board in the past 13 years.  In other words a 1 to 2 basis point variance in yearend NOL is normal.  And if there is a truly unforeseen event, then the premium option can be exercised.

But the average model does provide guidance to the fund’s most important month to month role of managing the investment portfolio.   As described in prior posts, the NCUSIF investment committee seems oblivious to the current period of historically low interest rates.

As recently as August during the increased market worry about inflation and rising rates, the  investment committee acted by rote. It laddered out $1.2 billion (6.2% of the fund) at an average weighted  life and yield of 5.7 years and .943% .  One seven-year investment decision alone cost credit unions $42 million in lost revenue because of the committee’s timing.

This is not the first time the committee has made similar decisions seemingly oblivious of the historically low interest rate period the economy is passing through.

These kinds of investment decisions, the continual piling on of expenses, and significantly erroneous estimates of loss are all examples of inadequate management.  No fund design can overcome human folly.

A Dynamic Model to Follow performance

Even with a 1.22% yield in 2021, the financial reports through September suggest the fund will in fact have positive net income and grow the equity ratio.  Why?  There is minimal loss reserve expense and net recoveries from prior loss reserves, a lower cost  of operations, and a decline in insured share growth from 2020.

A simple spread sheet model can track all these variables in real time.  The spread sheet accessed here uses the September numbers for the NCUSIF and insured share growth.  It projects a year end equity ratio of 28.19 which with the yearend 1% deposit true up equals an NOL of 1.2819%.

So even in an historically low yielding and poorly managed investment portfolio, the fund’s result is within the designed outcome.

Even Better News

Traditionally the fund’s equity has not included the loss reserve which has been expensed from earnings but not used.   When this reserve is added to the yearend actual equity ratio, then the NCUSIF reserves are even stronger, always exceeding .30, sometimes by as much as 10-20 basis points higher.

The Better Way design works, not because it is perfect, but because it is flexible and aligns resources with insured risk through constant 1% deposit required from credit unions.  It tales prudent management of expenses, careful loss control and conscientious investment management.

To continue NCUSIF’s performance pattern, four steps should be taken:

1. Audit and present the fund’s financial position following private GAAP accounting standards. Reset the NOL cap to 1.3% and follow accounting practice used prior to the 2001 change in recognizing the 1% deposit true up.
2. Reduce the OTR to 50% to correspond to the percentage of state-insured shares in the fund.
3. Update the oversight and transparency of the NCUSIF investments.
4. Continue to minimize losses to the fund by working through problem cases not cashing out losses problems by sales to outside bidders.

(Part 2 of 3 on  the  NCUSIF, A Better Way for credit union share insurance)

Throughout the NCUSIF redesign effort in 1984, credit unions had one primary concern with the new concept.  The question was, “If we send this money to Washington, how do we know government just won’t spend it?”

In response the legislation and implementation included multiple guardrails to prevent misuse of the fund by NCUA.  This included a legal cap of 1.3% on total fund size, after which a dividend must be paid; a limit on assessing premiums;  an annual independent  CPA audit following GAAP accounting; monthly public financial reports to the Board; the ability to withdraw the 1% deposit and a commitment to use the fund to minimize losses through 208 and other forms of temporary capital assistance.

Credit union concerns were well-founded.  The NCUA has become increasingly agile in charging the NCUSIF for its operating overhead expenses.

In the 13-year period 2008 through 2020, NCUA has spent a total of $3.321 bn.   Of that amount the operating fee has covered $1.172 or just 35% of NCUA’s expenditures. The NCUSIF has been charged $2.149 bn, or 65%.

Prior to the OTR change in 2001, the fund had never paid more than 50% of NCUA’s operations.

The Fund’s primary expense is for administration not insured losses

As a result, since 2008 NCUSIF’s operating costs exceeded the insurance loss provision expense of $1.880 billion by $270 million.  The primary financial role of the fund is to pay for insurance losses.  The cost have  been primarily through the Overhead Transfer Rate (OTR) process.  This process has been a source of ongoing  manipulation starting  in 2001.

The OTR History

Prior to 1985, the OTR was 33% which was based on the percentage of state-chartered insured shares in the NCUSIF.  In 1985 shortly after Ed Callahan left NCUA, the board raised the OTR to 50% even though the percentage of state chartered insured shares was still 33%.

This 50% ratio remained constant until 2001.  Dennis Dollar, as an NCUA board member and Chair (1997-2004),   made two changes in how the NCUA managed the NCUSIF. The first was making the OTR an annual adjustment, versus a fixed 50%  and secondly, delaying the recognition of the 1% deposit when calculating the yearend NOL.

The first change as reported in footnote 8 of the 2000 NCUSIF audit under the heading, Transactions with NCUA Operating Fund:

The allocation factor was 50% to the Fund and 50% to the NCUA Operating Fund for 2000 and 1999.  On November 16, 2000 the NCUA board voted to increase the allocation factor to the Fund for 2001 from 50% to 66.72%.

The only explanation for the change was that a new study of staff time spent on insurance versus supervision “indicated the rate should be changed to 66.72% for 2001.” ( Page 31 NCUA 2000 Annual Report)   This 33% increase in the transfer was at a time when state chartered  insured shares were only 44.6% of total NCUSIF insured risk.

Since this 2001 break with the long standing 50% transfer, the OTR has been recalculated every year reaching a peak of 73.1% in 2015.  The transfer rate for 2021 is 62.3 %.

The OTR has been the topic of contentious congressional hearings plus countless credit union objections to the arbitrary nature of the process.  See this link for congress’ questioning the OTR.  Even board members have expressed puzzlement with the explanations for how the transfer rate is determined.

Changing the NOL Calculation

The second change made in 2001 also had an ongoing consequence to the fund’s financial position and credit union’s dividend payment.

In 2000 the Normal Operating Level (NOL) was calculated as follows in footnote 5, Fund Capitalization:

The NCUA Board had determined the normal operating level to 1.33% as of December 31,2000, which considers an estimated $31.9 million in deposit adjustments to be billed to insured credit unions in 2001 based upon total insured shares s of December 31,2000. . . The CUMAA mandates the use of year-end reports of insured shares in the calculation of specified ratios and thus dividends related to 2000 will be declared and paid in 2001 based on insured shares as of December 31, 2000, as reported by the insured credit unions.

This was the method used in all the fund’s preceding years. Subsequently dividends of $99.5 million associated with insured shares as of December 31, 2000 were declared and paid in 2001.

But then this NOL calculation,  a consistent process  since 1985, was changed by the Board in 2001 as explained on page 21 if the Annual Report:

the deposit adjustment credit unions submit to the Share Insurance Fund to maintain their required deposit level of 1 percent of insured shares is not recorded until March 2002. In March, the equity ratio moved above the normal 1.3 percent operating level established by the Board. 

This delayed recognition allowed the NCUA to omit the required dividend in 2001 because the actual ratio exceeding 1.3% was not recognized until March.

Since the redesign of the NCUSIF’s in 1984, the adjustment to the 1% deposit had always been collected after yearend (as is the case today), but  still recognized in full when calculating the  December 31 NOL.

This change was discussed in the audited footnote 5 which revised the previous 2000 NOL calculation as follows:

The equity ratio at December 31, 2000, was 1.33%, which considered an estimated $31.9 million in deposit adjustments billed to insured credit unions in 2001 based upon total insured shares as of December 31, 2000. Subsequently, such deposit adjustments were excluded and the calculated equity ratio at December 31, 2000 was revised to 1.30%.

Thus the NCUA board gave itself the flexibility to delay recording the 1% true up versus continuing an  established and clearly understood ratio computation followed for almost 20 years in NCUSIF’s audited statements.

The Impact on the NCUSIF Financial Reports

The annual change in OTR from the long standing 50% and the delayed recognition of the 1% true up have given NCUA greater access to NCUSIF funding.

Since 2008, the operating expenses allocated to the NCUSIF have grown at an annual rate of 6.9%.   In contrast, following the change to a “variable” OTR, the operating fees paid by federal credit unions have increased only 3.6% demonstrating NCUA’s  increased use of the NCUSIF for its total budget.

The graph below shows NCUSIF operating expenses are almost all due to the Overhead Transfer with a different rate set each year.

Regaining Accountability for the NCUSIF

Credit unions were prudent to worry about NCUA’s instinct to spend money under its stewardship.  The NCUSIF is meant to be a financial safety net for the cooperative system. Instead it has been converted to a cash cow  anytime NCUA wants to reach in for more money via the allocation process.

Dividends that would have been paid under the 35- year long-standing 1.3% NOL cap were eliminated when  NCUA simply raised the cap to 1.39 in 2017 and 1.38 in subsequent years.  As many commentators pointed out, there was no objective  data supporting this change.

NCUA’s annual open-ended expense draw compromises the Fund’s primary role as a resource for insurance losses. Before the OTR change in 2001, the NCSIF paid six consecutive dividends. This kept NCUA’s commitment that credit unions could earn a dividend on their 1% investment in years of negligible losses.

Now NCUA takes their toll off the top rather than being limited to 50% a number which aligns with the proportion of state-chartered NCUSIF insured shares.

It is the NCUA board that made the above changes.  Accounting standards can accommodate different ways of presenting financial information.  These statements reflect management decisions, not accounting truths.  The changes in practice described about were not  in the best interests of credit union owners.  Auditors present, but do not approve such decisions.

NCUA’s 2022 budget is up for debate and comment.  The OTR should be set to reflect the proportion of state insured shares, not some opaque, manipulated internal study.  Expense control is a critical for  the proper management of any organization but especially a fund set up primarily to pay for insured losses.

Credit unions  have a duty to  speak up so the Board can right this series of decisions shortchanging the fund’s owners today.

Tomorrow in part 3 I will review the history of NCUSIF losses  to  analyze whether the long standing 1.3% cap and 1% deposit design provide sufficient financial resources for the fund.