Pension funds and their investment policy; a breach of trust or an opportunity?

Patterns of responsibility is the title of a book by Henk van Luijck, former professor of Nijenrode, on business ethics and corporate governance (ISBN 90 5261 237 4). In it he distinguishes two approaches to these themes; a compliance strategy and an integrity strategy. The first concerns an approach that only aims to comply with legislation and regulations. He shows that the focus on rules of the game not only leads to perverse behavior of the organization as a whole and the people who are part of it, but also to major inefficiencies, waste and injustice. We all know the examples.

Integrity strategy is based on the game goal. The object of the game is somewhat different from the outcome of the rules. Integrity strategy builds on the realization of the organizational goals based on shared beliefs, morality, the acceptance of responsibility and above all the willingness to account to each other and to the outside world, solicited and unsolicited.

Pension funds have a clear framework within which they report, among other things laid down in the PW and the current assessment framework (NFTK); The gamerules! This also applies to the accountability they give for their investment policy and the extent to which this serves the interests of the participants. The question now is how well they do that, where “good” is not determined by whether they comply with regulations, mainly by whether the rights and entitlements of participants have been optimized by the fund, which is in the missions of almost every fund. Hospitalized; the pin target!. In answering that question, I take the NFTK, i.e. steering on the funding ratio and the associated Asset Liability Management (ALM) as a starting point.

Telling about the investment policy is different from being accountable

Take any annual report of a pension fund and you will see that the accountability focuses on how the determination of the investment portfolio is set up, both in terms of process and organisation. The tasks, powers and responsibilities of the various fund bodies are discussed, followed by a description of the creation process and the presentation of the final result in this case strategic and/or tactical investment mix in EUR. All of this is put in the perspective of how the fund views developments in the financial markets. A classic example of compliance strategy; legislation and regulations are complied with.

It is not further explained whether the mix has indeed turned out to be optimal, given the risk appetite of the participants, the structure of the liabilities and developments in the financial markets. The returns achieved in the year are at best compared to benchmarks, but these are not interpreted in terms of, among other things:

  • Risks that were incurred but did not materialize (what part of the return was due to “gambling”)
  • The evolution of future risk and probability of benefits
  • The question of whether the financial risks incurred, regardless of whether they have become manifest, are in line with the risks that the fund intended to run on behalf of its participants

If funds were to do so, they would also provide insight into the impact on and the likelihood of the rights and entitlements of participants now and in the future. But they don’t. Yet pension funds start the process of arriving at an investment mix very thoroughly.

Achieving a sound investment mix is just the beginning

Every number of years, funds carry out a survey of the risk appetite among the participants. They derive a risk profile from this, which is subsequently translated into targets for return and risk, for example around the expected level of the funding ratio, the intended balance sheet risk (including bandwidth) or limits to the probability interval within which the funding ratio moves. And all this over a long time horizon, usually 15 to 20 years.

They then conduct an Asset Liability Management (ALM) study. Stochastic models are built around the dynamics of the participant population – death, disability but also wage development, retirement, entry and exit, etc. – and of course the financial markets, whereby various assumptions are made, including, for example, an interest rate view. Based on thousands, sometimes tens of thousands of (economic) scenarios, it is determined what the optimal investment mix is, given the targets set for return and risk over the chosen time horizon. This leads to a strategic asset allocation, also known as strategic weights.

The strategic weights lead to different mandates with various asset managers, including bandwidths within which the relevant asset class must remain. It is also these strategic weights that are controlled. When an allocation moves outside the stated bandwidths, the asset manager is told that he must either sell or attract assets. The basic principle is that the investment mix is optimal until the next ALM study or risk appetite study. In other words, in the best case scenario, the “fit” of the investment mix is targeted once a year. And that while reality is unfolding differently.

Administrative responsibility; take advantage of how reality unfolds

If investments and liabilities develop differently than anticipated in a period – and they do – this also leads to a different than expected development of the fund’s objectives (coverage ratio, return, risks, probability interval, etc.) over the chosen time horizon. You can distinguish 2 factors in this; variance and assumption change.

Variance is understood to mean the difference between the actual development during a certain period of the various variables containing the objectives of the fund (from funding ratio to probability interval) and the development as expected at the beginning of the year. Even if the development of reality fits in with the assumed underlying probability process that is below expectations, there will be variance. After all, sometimes you don’t roll the average with a clean, unweighted dice.

Variance can indeed be determined by doing another ALM study after a period, for example a month, with exactly the same model, the same parameters, assumptions, etc., but with the reality filled in for the past period instead of an expectation. Because pension funds limit themselves to managing mandates, they do not even determine the variance. For example, they cannot provide participants with insight into the impact of the past period on their entitlements and rights. In addition, they miss an opportunity to adjust the investment mix simply because they do not have the information. The excuse that the investment mix may not need to be adjusted at all makes no sense. An explicit statement that there is no better investment mix would be very valuable in the accountability to participants.

Assumption change
If it is concluded from the development of reality that the assumed underlying probability process is not correct, then this is a reason to adjust the (stochastic) models, parameters and scenarios. After all, the “system” of obligations and investments turns out to work differently than expected. The difference in quantities in which the fund’s objectives are contained in the original and the new model is called assumption change.

It goes without saying that to determine the assumption change, an ALM study must also be performed again; the results can then be compared with those of the old ALM. It is also evident that a different, optimal investment mix will then emerge.

The fact that pension funds do not frequently determine the variance and assumption change because they do not frequently perform an ALM does not only mean that they deny the participant insight into the impact on rights and entitlements, they also deny them the option of, among other things, higher entitlements and entitlements, lower risks or a narrower probability interval around the funding ratio.

Administrative integrity; prevent unnecessary loss of value by continuously doing an ALM

This is best described on the basis of case histories.

Dynamics of financial markets
Look at the interest rate declines in August 2019 when funds suddenly rushed into buying interest cover at high prices. From the 1st quarter of 2019, it was already visible that the funding ratio projections were shifting downwards, but that no fund on the retina that the plume of the probability interval not only moved along and, above all, became larger. This is a direct consequence of the fact that funds do not (have to) account for at least one year whether they have used the correct investment mix.

However, the dynamics in the financial markets are such that, within the chosen policy during the year, the investment mix (taking costs into account) must be aligned much more closely with the fund’s objectives. For example, interim ALM studies would have already made visible such a widening of the plume, as well as which shifts in the investment mix would have better safeguarded the fund’s objectives. Acting earlier means less loss of value and also knowing the instruments than a price advantage. And that can easily run into the hundreds of millions of euros.

Active investing vs passive investing
Many funds preach that benchmarks show that active investing does not outperform passive investing. Since the latter involves less investment costs, that would be preferable. By screening with benchmarks, only the return in the current year is looked at. This is simply a weak heuristic because they neglect the impact on the long-term development of the funding ratio, but (especially) also the development of the probability interval around this funding ratio. “Look how well we performed against the benchmarks?” But they have increased their risk profile and widened the plume.

Making a weighted statement about active versus passive investing is only possible if it is calibrated by means of ALM studies on the selected risk profile of the fund, or the objectives regarding return and risk over the chosen time horizon. In simple words, if you invest passively you must be able to demonstrate (calculate) that this is not canceled out by a much higher risk profile than is included in the objectives or a much higher risk profile with active investing.

It costs nothing!!!
The argument is often used that frequent ALM study costs far too much money.
Since models, parameters, interest rate vision, scenarios are simply present and computing power costs only a pittance these days, it must therefore lie in collecting and enriching the data.

Indeed, many pension funds have a complex, spaghetti-shaped information architecture with many decentralized databases at asset managers, custodians, risk engine providers, regulatory reporting providers, fiduciary managers, etc. And of course a multitude of vulnerable interfaces. Just “scrubbing” and putting data from all those different, outdated databases in one format, so that it is usable and available, requires a lot of effort and funds hardly do that. For example, pension funds hand over control of their data to all those outsourcing parties. And these parties have absolutely no interest in returning that control to the data owner, the pension fund.

So data is the problem… and the opportunity!

Complexity and costs are therefore a weak excuse. To start with the first; complexity arises because we first work from processes and only then from data. This way you get decentralized databases that you then have to rake back together. What if you first start with how you want to organize the relevant data and intelligence and only then look at what your organization, processes and systems will look like?

However, data is a very conceptual concept. That’s why it’s good to use a metaphor that everyone has a picture of: the BMW factory in Oxford! More than 300,000 different Minis are built there every year. There, the production process was not set up and then the supply lines of parts were organized, but first a thought was given to how those supply lines could be set up most efficiently, i.e. without stockpiling, in order to subsequently connect the production process to them. This led to cars being made for which the parts could be supplied. That seems inefficient, but it is actually not. Because there was no need to wait for parts, delivery times became shorter (less than 4 months). The production line never stops because a car under construction could be “parked” anywhere in the line until the required part was available. Once again it turned out that the number of parked cars remained very small, after all, the supply lines were central and there was rarely a shortage of parts. The analogy between car parts and data is obvious.

Current technologies make it possible to endlessly store data in any possible format (including interchangeably), label and classify it and then link it together at a fraction of the costs currently spent on information provision. to formulate. If a fund wants to adjust or expand its data model, this is not only possible at any time, the new data model can also be retroactively placed next to the old one. Data can be used to your heart’s content, both for the management of one’s own business operations and for the development of tactical and strategic objectives and the realization thereof.

Doing an ALM every day, if not near real time, literally takes a few minutes and a few cents. And if the board wants a different reporting format every day, this does not require endless, expensive coding, but can be done in fifteen minutes thanks to configuration.

Where pension funds in their current complex organization do not control their outsourcing landscape, taking control of data once again offers full control. After all you:

  • have synchronized, near real-time information
  • see the coherence across the organization
  • can make scenario analyzes & simulations regarding objectives at any time
  • can decide (much) faster and then act
  • can act (much) more efficiently/cheaper (it takes much less time)

Why then do pension funds stick so stubbornly to the approach of organizing processes, outsourcing them and then grabbing data again with great difficulty?

Restore trust through true accountability

In their annual reports, pension funds therefore say a lot about their investment and risk policy, but they do not show how their performance actually relates to their own risk profile and choices underlying their own ALM. They also do not show whether it could have been better based on those principles. In this way, a participant can never judge how well his pension fund is doing. Or is that precisely the intention….that it is not clear how the pension fund is doing?

The fact that pension funds do not even take themselves as benchmarks sounds like a politician who does not care about his own previous statements and positions. And that doesn’t earn you trust.

However, when a pension fund takes back control of its data, it can make it clear to its participants – every day if desired – for every event that it still uses the optimal investment mix or explain why it has implemented shifts in its mix. Not by protecting against benchmarks, but because it can relate its investment mix at any time and in any crisis to the risk appetite requested by the fund from the participants themselves. In fact, a fund can also show why participants would be worse off with a different investment mix. And that’s what restores trust.

Pension funds take your responsibility, take control of your data!

Let’s begin

Start creating more value by using data in your organisation with our guidence

Similar Posts