May 2007 Q.7

[Sifto]

This might be really obvious but what is the impact on the pension plan if a company is a non-taxable entity? I've worked with non-profit agencies but that particular status never impacted the funding of their pension plan (as far as i know anyway).

thanks.

[Kenny]

What do you mean? There is still the issue of deferred taxes for the participants. There is just no issue with a maximum tax deductible contribution from the sponsors perspective.

[Sifto]

sure but in this particular problem, the tax-status is not mentioned in the solution..is it a missed point or is it irrelevant?

[Kenny]

It wasn't missed. The answer specifically mentions that they assume the company is a not-for-profit and only discuss issues that are applicable to non-taxable entities. Notice in the response for company B the number of times the effect on the shareholder is addressed, but this is intentionally left out of the discussion of Company A. If you start talking about deductible contributions or a funding/investment strategy's effect on shareholders when discussing a not-for-profit then you are clearly not addressing the appropriate issues.

Tax status would not affect the general underlying actuarial funding principles, but second order effects on stakeholders would be different.

[Sifto]

I see..thanks...ok just to be sure..so for Company B which is a taxable entity, there is a tax advantage (amongst other items) to the s/holders if the company invests in bonds over equities. This doesn't hold for Company A since it's non-taxable. right?

[Kenny]

I see..thanks...ok just to be sure..so for Company B which is a taxable entity, there is a tax advantage (amongst other items) to the s/holders if the company invests in bonds over equities. This doesn't hold for Company A since it's non-taxable. right?Yes, assuming there is a difference between the tax rates. Based on the Gevrey/Vosne tax rates there is no tax advantage bc corporate tax rates = personal tax rates and equity tax rates = bond tax rates

[Sifto]

thanks Kenny.

[Jeremy Gold]

Be very careful with this question.

If the entity is not-for-profit, then it generally will not have s/h's and will not be maximizing value.

But if the entity is a corporation with a zero marginal tax bracket (for whatever reason), then the tax arbitrage is maximized for its s/h's. This occurs because the value of the tax arbitrage is proportional to (1-tau-sub-c) where tau-sub-c is the marginal tax rate of the corporation.

You can call me at my directory phone number or e-mail me. I would like to see a copy of Q.7 and the model solution if you send it to me.

Jeremy

[Kenny]

Be very careful with this question.

If the entity is not-for-profit, then it generally will not have s/h's and will not be maximizing value.

But if the entity is a corporation with a zero marginal tax bracket (for whatever reason), then the tax arbitrage is maximized for its s/h's. This occurs because the value of the tax arbitrage is proportional to (1-tau-sub-c) where tau-sub-c is the marginal tax rate of the corporation.

You can call me at my directory phone number or e-mail me. I would like to see a copy of Q.7 and the model solution if you send it to me.

Jeremy

Right, the issue whether shareholders exist or not.

Jeremy, you can get a copy of past exams and solutions here http://www.soa.org/education/resources/edu-multiple-choice-essay-examinations.aspx

Note that the Canadian and US question and answer are the same.

[fappakman]

You know its crunch time when the author of four of your articles posts on the message board.

[Jeremy Gold]

Thanks, Kenny, and good luck on the exams to you and the other AO posters.

I saw this question last year in the period after the exam and before final grading. I then asked for and was shown the grading outline. The grading outline is a much more detailed version of the model solution.

The description of Company A is not quite clear enough for me (and maybe some examinees last year) to know whether it is a not-for-profit or simply in the zero marginal tax bracket. The word "Company" raises doubts in both directions.

It is absolutely clear to me that the lower the tax bracket, the higher the arbitrage gain. Furthermore, deductibility is not the key. The key is that equities and bonds are equally taxed inside the tax shelter and equities are less taxed outside. This is why the arbitrage works even (at the Federal tax level) for pension plans sponsored by local governments and paid for by local taxpayers who also pay federal taxes.

I interceded to make sure that any examinee who understood Company A to have shareholders (and thus said it should invest in bonds for the arbitrage) would not be penalized in the grading.

I have no idea if anyone wrote the exam with this perspective and thus I don't know if any points were affected.

It appears that the model solution was not edited, as I would have liked, to cover the interpretation where Company A has shareholders.

[fappakman]

Good to know that the graders are given more than just the model solution! I don't think I knew that up until now. Thanks for the insight.

[Sifto]

thanks Jeremy. I see how "Non-taxable" entity could imply:

1) an entity with zero marginal tax or
2) a non-profit (and without s/holders as is mostly the case).

If 1) is the assumption then arbitrage is maximized by investing in bonds. If 2) is the assumption then there is no arbitrage.

[Jeremy Gold]

I guess, almost by definition, if there are no s/h's then there is no arbitrage.

But what is a not-for-profit's objective function? Is there a good reason for it to borrow to invest in the stock market? Is it self-funding or does it rely on donors? If it relies on donors, do they contribute appreciated stock (a great loophole for givers, BTW). I don't think the syllabus prepares one to address these issues and thus, frankly, I am not happy with the question at all.

I am also unhappy with the part of the answer that says:

"Will long term “cost” of plan allow for enough “time diversification” to let the equity investments pay off?"

I hope we are not teaching anywhere on the syllabus that time diversification is a respected idea in finance. And to include it in an answer to a question that claims:

"This is a synthesis question requiring candidates to describe the financial economics framework for pension plans."

I know that you all must be too stressed by the 2008 exam to care what I think about the weakness of a single (not to be repeated) question from the 2007 exam. So let me just repeat my best wishes.

[Sifto]

Thanks Jeremy. You're right in that the stress levels are through the roof right now - atleast for me. Let's leave this for now.

[Gord Ripley]

Jeremy,

The syllabus sure does teach us about time diversification and how it reduces equity exposure - it is one of the arguments in favor of traditional actuaries that we should spit out on the exam if asked. There is some stuff about how pension actuaries can predict the future with mean revision models too.

In one of your papers about public plans you mentioned that there was a tax arbitrage for taxpayers if gov pension assets are in bonds. I beleive that your point was that it is true to the extent that taxpayers are exposed to the pension risk. If the taxpayer is fully exposed then they would be just like shareholders, if they are not exposed then the arbitrage wouldn't apply since the taxpayer wouldn't adjust their own portfolio. I'm a bit fuzzy on this - is fully exposed like marked-to-market tax rates - am I on the right track? This is one case where even without shareholders - there is an arbitrage - isn't it?

[Jeremy Gold]

Gord,

Certainly marking tax rates to include the changes in pension surplus (deficit) immediately would be a sufficient condition for what Larry Bader and I call "exposure" in The Case Against Stock in Public Pension Plans. But weaker forms of exposure might also suffice in whole or in part. For example, suppose home prices adjusted to reflect delta surplus. More generally, if the taxpayers wealth correlated with the plan surplus, some degree of exposure might be recognized. If the change in taxpayer wealth were because of (rather than merely correlated with) changes in plan surplus, we would have a stronger form of exposure -- in that a change in pension asset allocation would change future taxpayer wealth. It seems obvious that eventually a higher equity exposure in the pension plan will affect the wealth of taxpayers for better or worse and with varying lags. If the lags are sufficently long, exposure (in the sense of the article) effectively disappears.

The whole arbitrage story is something of a metaphor used by economists as a proxy for a more general equilibrium pricing argument. If a firm lowers its pension equity allocation, no one expects individual s/h's to adjust their portfolios in contra-unison. But if the firm lowers equity allocation, and if the accounting were sufficiently transparent, the firm's beta should decrease and its lower expected future earnings would (ignoring tax effects) be offset (in the valuation of the firm) by the lower discount rate attributable to the lower beta.

If you don't mind, I would like a list of syllabus items that teach "us about time diversification and how it reduces equity exposure - it is one of the arguments in favor of traditional actuaries that we should spit out on the exam if asked. There is some stuff about how pension actuaries can predict the future with mean revision models too." If you are willing, you could post them or e-mail me at my yearbook address. Thank you and best of luck in the grading/waiting period.