Independent Report supports Colm Fagan's alternative AE proposal

Duke of Marmalade said:

Queries from the Council: An extensive and broad ranging discussion ensued. The Council members posed many questions, among those were:
• To what extent can we rely on past data for the future?
• The stock market cannot grow at a faster rate than the economy indefinitely –how is this being considered?
• To what extent can an assumption be made that it would be the State’s number one priority to bail out a pension fund in times of crisis? And should society bear those risks?
• Is there capacity in the market for everyone to do this?
• On what basis can we assume that this is a reliable equity basis?
• How was investment in index-linked bonds captured in CF analysis, in circumstances where high inflation caused bond values to fall significantly? (CF stated that he will revert to the Council on this.)
• What evidence do we have to satisfy ourselves that all associated risks can be managed?
• How has intergenerational fairness been taken into account?

Click to expand...

I came across a paper which may help the PA answer some of those questions: "Beyond the Status Quo: A Critical Assessment of Lifecycle Investment Advice" https://www.netspar.nl/assets/uploads/19.-Cederburg-ACO_Manuscript.pdf

"Our findings suggest that financial advice and pension regulations should be revised to consider all-equity strategies as viable and legal alternatives for retirement savers; we call for alternative approaches to mitigate the costs of short-term losses, such as financial education on staying the course, retirement account reporting standards that emphasize long-term performance, and regulations that assist retirement savers with maintaining a long-term focus."

nest egg said:

Let me try to put myself in their shoes. I suspect the PC (and DSP) are hesitant to recommend a novel approach for something like AE which will have such a broad impact. In addition, given the poor uptake in private pensions today, presumably 'any scheme is better than no scheme' is the prevailing mantra. Openly speaking, this isn't an unreasonable position to take.

I work for a large MNC and 'tried and tested' is often preferred to 'new and unknown'. However the best outcomes typically come from decisions taken after scrutinising several options.

We know that the smoothed approach has been subject to an independent evaluation. In the interest of fairness, has the proposed AE scheme had a similar independent evaluation?

Click to expand...

Slightly off topic but I think the take up rate at 14% of mostly modest salaries will be very low. The NEST experience is for a very large number of refusniks and their contribution rate is 8%

nest egg said:

I came across a paper which may help the PA answer some of those questions: "Beyond the Status Quo: A Critical Assessment of Lifecycle Investment Advice" https://www.netspar.nl/assets/uploads/19.-Cederburg-ACO_Manuscript.pdf

"Our findings suggest that financial advice and pension regulations should be revised to consider all-equity strategies as viable and legal alternatives for retirement savers; we call for alternative approaches to mitigate the costs of short-term losses, such as financial education on staying the course, retirement account reporting standards that emphasize long-term performance, and regulations that assist retirement savers with maintaining a long-term focus."

Click to expand...

Interesting, but I think naked exposure to the retirement timing risk of equity markets for someone funding for a pension of , say 15k, is debatable. I myself in my submission on AE back in 2018 suggested that there should be one fund for life but maybe invested 70/30 in equities/bonds.

Duke of Marmalade said:

Interesting, but I think naked exposure to the retirement timing risk of equity markets for someone funding for a pension of , say 15k, is debatable. I myself in my submission on AE back in 2018 suggested that there should be one fund for life but maybe invested 70/30 in equities/bonds.

Click to expand...

From what I've understood, that risk has been baked in, at least compared with a "Bal/I" (30% domestic / 30% international / 40% stock-bond mix). The ruin probability is still lower with the Stk/I (50% domestic / 50% international stock) approach.

Hi Brendan,

I'm not convinced by this report!! I suspect the Pensions Council would laugh at it!!

For example, take the 5 comparative scenarios - the first 5 columns above.

Columns 2&3 suggest a 60%/40% equity/bond split pre & post-retirement (now well known as sub-optimal)

Columns 4&5 go down the equity allocation as 120 less age (again now well known as sub-optimal)

This leaves us with column 1. (i) The glide path selected starts reducing the equity allocation well over 20 years from retirement (like come on) and (ii) lands at retirement with an equity exposure well shy of 50% (come on now again!) and (iii) even prior to the glide path starting only 90% are in equities. (iv) Even the equities throughout have clearly a much less favourable split in this approach than the ultimate comparator here (where the comparator is at 50% domestic/50% international). Basically, the Lifestyle approach chosen as the comparator has 4 readily identifiably weaknesses from my initial reading!

Have I convinced you?

Hi jas

I didn't get to read the report yet, just the summary.

I have marked it for detailed study on Monday.

Brendan

After dinner, by the fire with a large brandy!

jasdpace@gmail. said:

Hi Brendan,

I'm not convinced by this report!! I suspect the Pensions Council would laugh at it!!

For example, take the 5 comparative scenarios - the first 5 columns above.

Columns 2&3 suggest a 60%/40% equity/bond split pre & post-retirement (now well known as sub-optimal)

Columns 4&5 go down the equity allocation as 120 less age (again now well known as sub-optimal)

This leaves us with column 1. (i) The glide path selected starts reducing the equity allocation well over 20 years from retirement (like come on) and (ii) lands at retirement with an equity exposure well shy of 50% (come on now again!) and (iii) even prior to the glide path starting only 90% are in equities. (iv) Even the equities throughout have clearly a much less favourable split in this approach than the ultimate comparator here (where the comparator is at 50% domestic/50% international). Basically, the Lifestyle approach chosen as the comparator has 4 readily identifiably weaknesses from my initial reading!

Have I convinced you?

Click to expand...

Your post does highlight a question, what will be the default approach adopted for the AE scheme?

It seems only with that information could the PC evaluate its merits relative to any other approach.

Fair point, Nest Egg - I agree with you. I was just commenting on that report on its standalone merits and based on its intended audience.

nest egg said:

Your post does highlight a question, what will be the default approach adopted for the AE scheme?

Click to expand...

The draft legislation states that the default strategy must take age into account etc. They are clearly thinking NEST, c. 50 retirement age based funds.

jasdpace@gmail. said:

Columns 2&3 suggest a 60%/40% equity/bond split pre & post-retirement (now well known as sub-optimal)

Columns 4&5 go down the equity allocation as 120 less age (again now well known as sub-optimal)

Click to expand...

jasdpace@gmail. said:

Have I convinced you?

Click to expand...

Hi jad

I had a quick look at that paper.

Is your argument that they are unfairly comparing a 100% stocks strategy with a life cycle strategy that no one uses anymore?

From memory, and Duke will correct me if I am wrong, the original straw man proposal assumed a life cycle strategy. It may be well known to be sub-optimal, but the guys behind the AE scheme did not know. I don't know what the current proposal for the default investment plan is.

We also know that most people just accept the default investment plan, whether it is sub-optimal or not.

Most of the life cycle strategies that I have seen assumed that people switched from equities to 100% bonds or cash on retirement.

You and I might know that this would be sub-optimal, but I think it's still a widespread practice.

So I wonder did the authors compare their 100% stock allocation strategy with what is actually happening in practice in America?

Most retired people I know won't look at equities. They want their money in cash as they think it's "safe".

Although these strategies are well known to be sub-optimal, they are in widespread use:

The first strategy shown in Panel A of TableIfollows the advertised unconditional glidepath from a TDF offered by a major investment firm.

....

The “120−Age” rule shares its theoretical underpinnings with the TDF strategy, but it represents a somewhat simpler investment
heuristic that is popular among financial advisors. In Choi’s (2022) survey of the 50 most popular books on personal finance, for example, he finds that nine explicitly recommend an asset allocation that is a decreasing linear function of investor age.

So they seem like fair strategies to compare with.

If nothing else, the paper confirms that a 100% stocks allocation is better than these two strategies which are widely know to be sub-optimal.

New Ireland promotes such a fund for pensions savers.


At retirement, 75% is in "Prime 3" which is about 25% in equities.

I presume that the other pension companies promote similar products.

They don't say anywhere prominently that this approach is now known to be sub-optimal.

Brendan

Brendan Burgess said:

From memory, and Duke will correct me if I am wrong, the original straw man proposal assumed a life cycle strategy.

Click to expand...

It used the Pensions Authority's pension calculator. I think that has lifestyling but crucially it assumes a "conversion factor" at retirement, which I presume means an annuity.
I got round to reading the paper linked by @nest egg. I have a couple of issues with its conclusions. It concludes that for the US saver, 100% in US stocks does not work but that a 50/50 mix of US and other developed stockmarkets beats the pants off all other strategies. It puts this down to diversification. But that is a misrepresentation; no-one claims that diversification systematically increases returns, it reduces volatility. The methodology was to "bootstrap" from historic returns so really all that is being demonstrated is that over the last 100 years or so the other markets have been catching up with the US and in doing so performed considerably better. In fact this is another major criticism of the methodology. It states that it uses 1,000,000 simulations which might suggest great precision. But since it is bootstrapping from the past it is merely reproducing past performance. The last 100 years have been one of unprecedented technological advances in productivity. With significant headwinds (climate change) in the future this looks overly optimistic.

I suppose what he is arguing is that over the long term the performance of a particular stockmarket is a random variable of that particular economy. Certainly, for Ireland the case for geographical diversification is compelling. But in US vs The Rest we are really comparing two equals. The US has done considerable worse than The Rest over the last 100 years but which has been the "lucky" one? The authors are assuming that the average of two outcomes is a correct predictor of the future - an heroic assumption.

And finally a wonkish point. By bootstrapping 1,000,000 times the Central Limit Theorem dictates that really only two moments matter - the mean and variance. They might just as well have used a simple GBM model with the mean and variance of the past. But "bootstrapping" from actual history seems so much more representative. A false argument is that bootstrapping from history picks up the possibility of events like the Wall Street crash or Black Monday. Not so, these are "black swan events" under GBM i.e. no way occurring several times in a century.

Duke of Marmalade said:

And finally a wonkish point. By bootstrapping 1,000,000 times the Central Limit Theorem dictates that really only two moments matter - the mean and variance. They might just as well have used a simple GBM model with the mean and variance of the past. But "bootstrapping" from actual history seems so much more representative. A false argument is that bootstrapping from history picks up the possibility of events like the Wall Street crash or Black Monday. Not so, these are "black swan events" under GBM i.e. no way occurring several times in a century.

Click to expand...

My bad. This wonkish point applies to using bootstrapping say (over 3,000) daily returns to produce 10 year illustrations. The bootstrapping in the paper uses 10 year blocks so hardly any impact from the CLT.

Duke of Marmalade said:

Interesting FT article damning Lifestyling

Click to expand...

There's an archived version too for anyone who can't access Duke's link - https://archive.ph/3MaOQ

I read about these AE proposals with interest. I think the past 12 months have thrown the traditional idea of de-risking by switching towards bonds as you age into disarray.

Any arguments decision makers have that it shouldn't be done because nowhere else has done it before should be completely ignored and the thousands that saw their pension funds drop dramatically with bond price collapses last year will attest to the tried and tested methods perhaps not being the best methods.

Bonds have come back somewhat since the trough but I imagine many received a shock upon receiving annual statements, panicked being so close to retirement and may have moved out of the "risky" bonds whilst they were at their lowest.

Even forgetting this bond collapse, they generally offer less returns and the increase in life expectancy means that someone facing the average number of years they now face in retirement should understand the risk inherent in bonds over such a long term investment.

Personally, in retirement, if my pension fund resulted in an income that was VERY tight then, and only then, would I look at an annuity to bring in some sort of guarantee and alleviate most worries. If a 5% withdrawal rate, even after a 30% price drop, still resulted in reasonable comfort, I'd be 100% equities. A 5% withdrawal is high but not if it's after a 30% plummet in equities; which history tells us will recover. A bond allocation would never come into the picture for me personally.

@ronaldo Your last paragraph describes the issue well. In general the target audience for AE would not be “comfortable” after a 30% drop in income in retirement. It is difficult to avoid concluding that they have no choice but to pay the very high price of “insuring” their income through an annuity or similar. Colm’s proposal purports to remove the huge cost of this insurance by pooling the volatility risks between cohorts of retirees.

@Colm Fagan

Did you ever hear anything more on anything related to your alternative proposal? Is the legislation/plan so far advanced that they cannot pivot if they wanted to?