Why Anthony Hilton is Wrong on Pensions 

“Expected returns don’t pay benefits, cash does”

Anthony Hilton recently wrote a stinging attack on pension consultants -like myself- who advise defined benefit pension schemes to measure, and hedge, their liabilities with reference to gilt yields.

Here’s one (of several) reasons why he’s wrong.

His argument – that you could fund based on the much higher expected returns on risky investments – might perhaps be justifiable in a world where all corporate sponsors were rock solid and would last forever (clearly we do not live in this world – and even then it would expose companies to some needless nasty surprises along the way, but anyway).

However this completely misses the point that a major reason we fund pensions at all is precisely to provide security in a situation where the corporate sponsor ceases to be able to make payments itself. The reason we need to measure deficits is to get a picture of how secure the benefits might be in the absence of the employer, and to take corrective action (eg topping up contributions) if the situation is off-track, before it is too late.

And in those situations of sponsor company failure, we would hit a major snag under Mr Hilton’s approach: it’s hard cash that must be used to secure the benefits – Mr Hilton’s expected returns won’t cut it I’m afraid. Just ask the pensioners of BHS scheme, or indeed the allied steel and wire groups, still campaigning for their pensions over a decade later. This second example pre-dates the Pension Protection fund, so thankfully pensions today are better protected, but the conclusion for scheme funding – that you can’t rely on high future expected returns to discount liabilities – remains valid.

Pensions need to be paid to members in real cash, and it flies in the face of both accepted theory, and common sense, that the amount of money needed to provide these benefits can be reduced depending on the assets held to deliver them.

Today’s unfortunate reality is that the defined benefit system in the UK is on average chronically under funded compared to the benefits it has promised. Time and effort would be far better spent on considering the tough choices that might need to be taken, rather than on attempts to deny the existence of a problem in the first place.

Unfortunately Mr Hilton’s ill-judged remarks from an otherwise respected journalist damage the hard work that many of us in the industry have been doing for many years to try and secure the benefits and financial futures of those members dependent on defined benefit pensions for their retirement.

Grim Summer for DB Pensions

It’s not been a good summer for the financial health of UK DB pension funds.

Data released by the PPF in September underlined a summer of bad news for schemes and scheme sponsors.

A continued trend of falling gilt yields have seen liability values increase dramatically – adding more than £200bn to the aggregate deficit compared to the start of the year. The aggregate funding level (on the PPF basis) fell to 76% at the end of August, the lowest value in the 10-year history of the series. For context the funding level on this basis was close to 100% at the end of 2013.

Long-dated gilt yields have fallen from around 2.5% at the start of the year to around 1.2% by the end of August.

The result of the referendum vote certainly added impetus to the move lower in gilt yields, as “lower for longer” became a more likely scenario, and this was re-enforced by the BoE’s announcement of renewed bond purchases (quantitative easing) and a cut in interest rates to 0.25%.

While asset markets have generally performed positively – especially overseas equities in £ terms, these good results have been insufficient to keep pace with unhedged liabilities.

It seems likely that actuarial valuations as at 30 June or 30 September are likely to contain bad news for corporate sponsors, prompting tough conversations such as those happening at plastics manufacturer Carclo, and negative headlines such as those at Associated British Foods. These are likely to become more and more common as we move forward, as highlighted by LCP in their Accounting for Pensions report.



Black Box Thinking

The best book I read over the summer was Matthew Syed’s Black Box Thinking.

The central theme of the book is really fear of, and reaction to, failure.
We have an allergic aversion to failure. We try to avoid it, cover it up and airbrush. The phenomenon of cognitive dissonance  is the name for the deeply-rooted behavioural trait that causes us to naturally reject ideas or even evidence that conflicts with our own worldview. This can be incredibly damaging to progress in many cases.
There’s a huge need to learn from failure, it can be extremely helpful (Syed cites the example of the aviation industry learning from air disasters to vastly improve the safety record).
Readers of similarly themed books (eg the work of Charles Duhigg, Khoi Tu or even David Eagleman)  will find a lot of the examples used by Syed a little tired and overdone by now. However I found that Syed was able to extract sufficient new insight from some of these well trodden case studies and weave them together with the central theme effectively.
Creating  by experimenting is often more effective than creating by blueprint.

Cognitive dissonance / confirmation bias

There are some powerful behavioural psychological forces at play that can be quite counter-productive to progress in today’s world. For example, the tendancy of reframing when faced with evidence that we’re wrong … divorces us from the pain of recognising that we were wrong. It’s not even conscious when it happens.
For example –

Open vs closed loops.

Syed defines open loops as operating by benefiting from feedback for example aircraft black box & medical randomised control trials.
Closed loops do not systematically collect feedback AND more seriously do not have the mindset to confront, recognise and learn from failure. Closed loops systems are dangerous because they  block progress – whether that be progress in safety, improving care, innovation or surviving in the commercial world.

Evolution itself is the best example of learning from failure.

Narrative fallacy Vs RCT

Narrative fallacies arise inevitably from our continuous attempt (need) to make sense of the world around us. The explanatory stories that people find compelling are simple and concrete. But they often assign a greater role to talent/stupidity/intentions than to luck and often rely on a few events that did happen rather than the countless that didn’t.
Stories are good but beware the narrative fallacy. Eg scared straight. Statistical biases.
Need counter factual and control group.

Marginal gains & feedback loop

Marginal gains is not about making small changes and hoping they fly. Rather it is about breaking down a big problem into small parts in order to rigorously establish what works and what doesn’t.
Break a performance into a component parts & you can build back up with confidence > brailsford
Some programs are hard to create controlled trials for, eg aid to Africa. Break down into component parts >> marginal gains
The existence of a local maximum reveals the inherent limitation of marginal gains. Sometimes you need a big leap forward to get past a local maximum. Need to do both marginal gains and big-picture thinking.


Contradictory information jars us psychologicaly. It nudges us into looking for unusual connections. Innovation comes from making new connections between familiar things.
Find a hidden connection to solve a problem. Failure and epiphany are linked. Brilliant ideas can emerge from engagement with a problem for months or years.
Innovation is context dependent – a response to a particular problem at a particular time & place.
Big picture & small picture. Innovation + discipline = success. There exists a threshold level of innovation required for a firm to be successful, beyond that it depends on the discipline to implement.

Book review – Smarter, Faster, Better 

After really enjoying Charles Duhigg‘s excellent first book The Power of Habit, it was an easy choice  for me to grab a copy of his second book on the way to my holiday, having been given a recommendation from Mitesh. I wasn’t disappointed!

Essentially the book looks to explore ways in which we as individuals and teams can be more creative and productive. Some of the points and ideas are familiar, but I really like Duhigg’s style of weaving neatly summarised academic literature into memorable real world stories and characters.

Here’s my top 4 takeaways-
1. Single most important tip for teams: build a commitment culture through fostering psychological safety.

A commitment culture (as opposed to a star culture) is one where the whole team is genuinely committed to helping each other reach a common goal. Psychological safety means that each person on the team can speak up, contribute ideas, and critique ideas. Everyone gets their turn to speak and meetings are not dominated by a couple of individuals.

2. Motivate yourself and others by making choices that put you in control.

It turns out that the need for control is pretty fundamentally hard wired into us from an early age (I don’t have kids, but my friends that do tell me this is something they experience frequently from the age of about 2). A perceived lack of control over a difficult of demanding task can be stifling for our motivation (just ask any student approaching revision for exams!) But as adults we can use this to our advantage. For ourselves, approaching draining or difficult tasks can be easier if we start by framing a choice (choose the location for a difficult meeting, taking control of your availability). When managing others it can be incredibly powerful and motivating to pass control to them – allow them to take key decisions relating to the project. This also ties in with agile principles.

3. Build mental models

Our brains are set up to build models of the world around us and constantly evaluate information received against the model (David Eagleman writes more on this in his excellent book on neuroscience, incognito). We can harness this in a working environment by constantly building a model of how we expect a given day, interaction, meeting or project to play out. Evaluating what happens in reality relative to this model can help better decision making. In the book, Charles Duhigg uses some excellent contrasting examples of aviation incidents to really bring this home.

4. Make data disfluent, in order to understand it better

We live in a world that has never been richer in terms of data. We each generate huge amounts of data everyday and carry in our pockets devices capable of processing data that previous generations couldn’t dream of. But how do we turn that data into actual information and insights?

Paradoxically a great way of doing this is often to go back to basics. Make the data harder to interact with at first. Draw graphs by hand, write datapoints out longhand on flash cards. By doing this we are forced to interact with the data more. We build theories about what the data contains, and in testing these theories we learn the important lessons. This is also why taking handwritten notes can be more powerful than typing notes, precisely because it is MORE labour intensive.

There’s plenty more in this very readable book, granted not all the concepts are revolutionary, but I would be surprised if you couldn’t find a few real actionable insights to take away and apply day to day. I certainly did and I look forward to implementing these with my team.

Pension Funding Levels Plunge in January 2016

The first 2016 update of the PPF index (of UK defined benefit pension funds) is just out, and it wasn’t pretty

Falling long-dated interest rates over the month (the 30 year gilt rate fell by 0.3% to 2.3%) caused the present value of the liabilities to increase by some 6%, while the assets were virtually unchanged (presumably some positive returns from bonds being offset by negative returns from equities).

Overall this caused a fall in the aggregate funding level of 4.4%. This reverses the gains made in the 4th quarter of 2015. Only twice in the 10-year history has the funding level been lower – in January last year and in May of 2012 when the funding level fell below 80%.

Worth noting is that the funding ratio tracked here is in respect of the PPF-liabilities, which are usually less than the standard contractual liabilities of the scheme. Typically the self-sufficiency liability valuations that scheme trustees might use would be 25% higher.

The total deficit increased by nearly 80bn to the second-highest on record at 304bn.

I estimate that this increase in deficit would add roughly 0.4-0.5% p.a. to the returns required every year to become fully funded over the next 15 or 20 years (which might not sound like a lot, but given the difficulty and risk associated with achieving the already high returns many schemes need this can be quite challenging). Put another way, this extends the time to full funding for a scheme which resembles this average position by around 3 years.

The variation in the overall funding position remains driven by the liabilities, which have had a volatility nearly three times that of the assets over the life of the data series, pointing toward LDI strategies as the first port-of-call for pension schemes trying to stabilize their funding position.

In a paper written last year I commented on the asset allocation changes in UK pension schemes. There has been some adoption of LDI strategies, but much de-risking has consisted of removing return seeking assets without fully hedging LDI risks, which leaves schemes vulnerable to interest rates falling a we have seen this month.

PPF Jan 2016