Direct and non-direct recognition loans: how borrowing affects dividends

When you borrow against a whole life policy, two things happen at once: you start owing interest on the loan, and the insurer has to decide what your borrowed cash value earns while it is pledged. A company that sets the dividend as if the loan did not exist is called a non-direct recognition insurer. A company that adjusts the dividend on the borrowed part is called direct recognition. Neither label tells you which is cheaper for you. The cost depends on the rates involved in the particular year, and on what the contract and the dividend notice actually say.

Start with what a dividend is

A participating whole life policy pays a policy dividend when the insurer's experience is better than the pricing assumptions it used. Those assumptions covered three things: investment earnings, the cost of mortality claims and the expense of running the company. When earnings, claims and expenses come in better than planned, the surplus is shared with policyholders through the dividend scale, which the company's board sets, usually once a year. Dividends are not guaranteed, and the amount can fall as well as rise. The guide on whole life dividends: how participating policies pay covers the full mechanism, and how a whole life dividend travels from company earnings to your statement follows one dividend from the investment portfolio to your annual statement.

For this post the important piece is the interest component. A dividend is built partly from what the insurer earns on the assets backing your policy's reserves and cash value. That earning is a rate, often called the dividend interest rate, and it is applied to a policy's reserve or cash value. Whether the rate applies equally to borrowed and unborrowed amounts is exactly the direct versus non-direct recognition question.

What happens to your cash value when you borrow

A policy loan does not take money out of your cash value. The insurer pays you out of its general funds and holds your cash value as collateral. On paper the cash value stays in the policy, and the loan sits beside it as a debt. That is why the policy stays in force and why the death benefit is reduced only by the unpaid loan, not by the full borrowed amount.

From the insurer's side, however, the loan changes the economics. Before the loan, the assets behind your cash value were invested in bonds, mortgages and other holdings earning the portfolio rate. After the loan, part of those funds has been paid out to you, and in exchange the insurer receives loan interest at the contract's loan rate. If the loan rate is higher than the portfolio rate, the insurer earns more on that money than it did before. If the loan rate is lower, it earns less. A direct-recognition company says: that difference belongs in the dividend, so we will credit the loaned portion differently. A non-direct company says: we will smooth the effect across all policyholders and set the dividend as though every policy were unborrowed.

Direct recognition, in plain terms

Under direct recognition, the dividend interest rate applied to the loaned portion of the cash value can differ from the rate applied to the rest. Often the company ties the loaned-value rate to the loan rate in some fashion. Depending on the design and the year, the loaned portion can earn less than the unloaned portion, or it can earn more.

A few features recur:

• A loan-linked credit. The loaned portion might be credited at a rate related to the loan interest rate or to a published index, rather than at the full dividend interest rate.
• A variable loan rate. Many direct-recognition contracts pair the approach with a loan rate that moves with a published index, so the pass-through works in both directions.
• Timing rules. The adjustment is typically based on the loan balance on a particular date or averaged over the year, so when in the policy year you borrow can matter.

Non-direct recognition, in plain terms

Under non-direct recognition, the dividend is declared as though there were no loan. A policy with a $30,000 loan receives the same dividend as an identical policy with no loan. The borrower pays the loan interest, but the dividend itself is not reduced because of the borrowing.

Typical features:

• A stated, often fixed, loan rate. The company sets a loan rate in the contract and takes the risk that its own earnings differ.
• Cost shared across the pool. If many policyholders borrow when the loan rate is below the portfolio yield, the company's earnings are lower, and that shortfall is a cost to the whole participating pool, not just the borrowers. In the other direction, borrowers are charged more than the portfolio earns.
• Simplicity. The dividend notice does not carry a separate loaned and unloaned calculation.

Each design has supporters. Advocates of direct recognition say it treats borrowers and non-borrowers alike by matching what each group costs the company. Advocates of non-direct recognition say it is simpler and more predictable for the borrower. Both are legitimate, regulated approaches, and neither guarantees a particular outcome.

A worked comparison with invented numbers

All figures below are hypothetical. They are not dividend rates, loan rates or results of any company, and the actual dividend has other parts that this example leaves out.

Say a policy has $100,000 of cash value. The owner has a $30,000 loan at a 6% loan rate, so loan interest is $30,000 x 0.06 = $1,800 a year. Suppose the interest-related part of the dividend would be 5% on unborrowed cash value. For simplicity, ignore the mortality and expense parts of the dividend and the growth in cash value during the year.

Non-direct recognition. The dividend interest component is 5% x $100,000 = $5,000. The loan did not change it. The policy's net cost of the loan is the interest paid, $1,800, which is 6.0% of the $30,000 borrowed.

Direct recognition, loaned portion credited less. Suppose the loaned $30,000 is credited at 3.5% and the unborrowed $70,000 at 5%.

• $70,000 x 0.05 = $3,500
• $30,000 x 0.035 = $1,050
• Dividend interest component = $4,550, which is $450 less than $5,000

The owner pays $1,800 in loan interest and gives up $450 of dividend, so the all-in cost is $2,250, or 7.5% of $30,000.

Direct recognition, loaned portion credited more. Suppose rates have risen and the loaned $30,000 is now credited at 6.5%.

• $70,000 x 0.05 = $3,500
• $30,000 x 0.065 = $1,950
• Dividend interest component = $5,450, which is $450 more than $5,000

The owner pays $1,800 and gains $450 of dividend, so the all-in cost is $1,350, or 4.5% of $30,000.

The lesson is that the stated 6% loan rate is only part of the cost. The true net cost is the loan interest plus or minus whatever the loan does to your dividend. A direct-recognition policy can cost more or less than a non-direct one in any given year. That is why comparing two policies by loan rate alone, or by an agent's remark that a company is "non-direct," can mislead. The structure of the credit matters, and so does the direction of interest rates.

Where other policy types fit

The same question exists outside participating whole life, under different names. Universal life and indexed universal life policies may credit a lower rate, or a different index treatment, to borrowed values. Some have a "participating loan" feature, others a "wash" feature in which loaned value is credited at about the loan rate. Those variations are the subject of fixed, variable and wash loans compared: what borrowing costs net. The common thread is that a policy's cash value is not always earning the same thing before and after you borrow, and the contract is where to find out.

Why this matters for borrowing strategies

Some financial planners and promoters present a policy loan as nearly free because "the cash value keeps growing while you borrow." That statement is true in a narrow sense under non-direct recognition, where the dividend is set without regard to the loan, and it can be partly true under direct recognition in a rising-rate year. It is not automatically true. The guide on infinite banking with whole life: how it works and the catch covers the case for and against borrowing from a policy as a financing method, and notes that the net cost is the real measure. The fair way to evaluate any such claim is to ask for the net cost in a stated year, and to see the dividend notice from a year when you actually had a loan.

Two further points apply either way:

1. Dividends are not guaranteed. A direct-recognition formula can look attractive in an illustration that assumes today's scale. If the scale falls, see what happens in what to do when your whole life dividend scale is lowered.
2. Compounding loan interest. If you do not pay the interest, it is added to the loan and the dividend adjustment applies to a growing base. policy loan worksheet: interest, repayments and what the policy can carry shows the growth by year.

Four misunderstandings worth clearing up

"Non-direct means the loan is free." It does not. You pay the loan interest in full, and the loaned cash value is still collateral. What non-direct means is only that the dividend itself is not changed by your borrowing. In the invented example above, the non-direct owner still paid $1,800.

"Direct recognition always punishes borrowers." It does not. The adjustment follows the credit assigned to the loaned portion, and that credit can sit above or below the rate on unborrowed value. In the example the same owner came out $450 worse in one year and $450 better in another.

"The cash value on my statement is untouched by a loan." The statement may show cash value and loan separately and look fine, while the dividend line quietly reflects an adjustment. If you borrowed and the dividend fell more than the dividend scale announcement suggested, ask whether recognition is the reason.

"A company's method never changes." The dividend scale is set annually, and a company may change how it sets rates on loaned values within what the contract allows. The contract's loan provisions limit the loan rate, but the dividend is a non-guaranteed element. An old brochure is not a promise.

Reading your own statement in three steps

If you already have a loan, you can often check how it is being treated without asking anyone first.

1. Find the loan section of the annual statement: balance at the start of the year, interest charged, payments, balance at the end.
2. Find the dividend section. Look for a line that separates dividends on loaned and unloaned values, or an explanatory note. If there is only one number and no note, the company is probably non-direct, or it does not show the breakdown.
3. Compare to the prior year if you had no loan then. A noticeable drop in the dividend that is not matched by a lower scale is a prompt to ask.

If the statement is silent, a written question to the insurer, referencing the policy number and the year, is a fair request.

What to ask the insurer

You do not need to learn the company's formulas. You need straight answers to a short set of questions, ideally in writing:

1. Does the company use direct or non-direct recognition, and is that stated in my contract or only in the dividend notice? The contract wording carries more weight than a brochure.
2. If direct, how is the rate on the loaned portion set? Is it tied to the loan rate, an index, or a decision by the board each year?
3. What dividend interest rate was credited last year to the loaned portion, and to the unloaned portion? The two figures side by side tell you what the formula did in practice.
4. Is my loan rate fixed or variable, and what is the maximum? A cap limits one side of the equation.
5. How does the timing of my loan during the policy year affect the adjustment? A loan taken just before an anniversary may be treated differently from one taken just after.
6. What does the in-force illustration assume for loaned values? Ask for an how to request and read an in-force illustration that shows a loan, and check whether it uses current or lower dividend assumptions.
7. Can I see a year's dividend notice that shows both amounts? The notice is the evidence.

If the answers are vague, that tells you something. The state insurance department is the right place to ask when a company cannot or will not explain how it treats borrowed values.

A short decision guide

• If predictability matters most, and you are comfortable with a fixed, stated loan rate, a non-direct design may match your preference, as long as you also check the loan rate level.
• If you expect to borrow in a period of high interest rates and the contract passes those rates through to loaned values, a direct design could be cheaper in those years.
• If you are not sure when you will borrow, or at what rate environment, focus less on the label and more on the worst case: the highest loan rate the contract permits, and the lowest credit on loaned values.

None of these is a recommendation for a company or product. A fee-only planner or an insurance professional who is not paid on the sale can walk through your own contract.

Key takeaways

• Direct recognition lets the insurer adjust the dividend on borrowed cash value; non-direct sets the dividend as if there were no loan.
• Direct recognition can raise or lower the net cost of a loan depending on the credit assigned to the loaned portion in a given year.
• In the invented example, a 6% loan costs 6.0%, 7.5% or 4.5% net, depending on the treatment, so the loan rate alone is not the cost.
• Dividends are not guaranteed, so test any borrowing plan against a lower dividend scale.
• Ask for the dividend interest rates credited to loaned and unloaned values in the same year, and get the contract wording in writing.