Category: investment

One minute guide to real-world AI implementation 

McKinsey just published an excellent and comprehensive paper covering how Artificial Intelligence (AI) can deliver real value for business.


The only issue – at 80 pages it’s a lot to read.

A lot of the use cases focus on retail, energy and education, one angle I find particularly are the read-across of these examples into service based and business-to-business environments. There are definitely some relevant points that could map to a services/B2B worlds: for example the automation of admin tasks for teaches, more targeted sales and marketing and more personalised customer service.

Here’s my take on the key points from the document:

1. No shortcuts: first data & digital, then AI

AI becomes impactful when it has access to large amounts of high-quality data and is integrated into automated work processes. AI is not a shortcut to these digital foundations. Rather, it is a powerful extension of them.

The firs thing firm’s need to do is come up with a real business case for AI that relates to the firm’s strategy, this requires separating the hype and buzz around AI from its actual capabilities in a specific, real-world context. It includes a realistic view of AI’s capabilities and an honest accounting of its limitations, which requires at least a high-level grasp of how AI works and how it differs from conventional technological approaches.
Each new generation of tech builds on the previous one – this suggests AI can deliver significant competitive advantages, but only for firms that are fully committed to it. Take any ingredient away—a strong digital starting point, serious adoption of AI, or a proactive strategic posture—and profit margins are much less impressive. This is consistent with McKinsey findings in the broader digital space.
Technology is a tool and in itself does not deliver competitiveness improvements.

2. Areas to focus on to create real value: project, produce, promote or provide 

To fulfil the expectations being heaped upon it, AI will need to deliver economic applications that significantly reduce costs, increase revenue, and enhance asset utilization.

Mckinsey categorized the ways in which AI can create value in four areas:(1) enabling companies to better project and forecast to anticipate demand, optimize R&D, and improve sourcing; (2) increasing companies’ ability to produce goods and services at lower cost and higher quality; (3) helping promote offerings at the right price, with the right message, and to the right target customers; and (4) allowing them to provide rich, personal, and convenient
user experiences

3. Data ecosystem & staff culture to the fore 

Firms must conduct sensible analysis of what the most valuable AI use cases are. They should also build out the supporting digital assets and capabilities. Indeed, the core elements of a successful AI transformation are the same as those for data and analytics generally. This includes building the data ecosystem, adopting the right techniques and tools, integrating technology into workplace processes, and adopting an open, collaborative culture while reskilling the workforce

4. Take a portfolio approach focused on use cases in short, medium and long term, be lean, fail fast & learn

A portfolio-based approach to AI adoption cases, looking at use cases over a one- to five-year horizon, can be helpful.

In the immediate future, McKinsey suggest a focus on use cases where there are proven technology solutions today that can be adopted at scale, such as robotic process automation and some applications of machine learning. Further out, identify use cases where a technology is emerging but not yet proven at scale. Over the longer term, McKinsey’s view is to pick one or two high-impact but unproven use cases and partner with academia or other third parties to innovate, gaining a potential first-mover advantage in the future. Across all horizons, a “test and learn” approach can help validate the business case, conducting time-limited experiments to see what really works and then scaling up successes. Fast, agile approaches are important.

5. Don’t be a hammer in search of a nail … 

To ensure a focus on the most valuable use cases, AI initiatives should be assessed and co-led by both business and technical leaders. Given the significant advancements in AI technologies in recent years, there is a tendency to compartmentalize accountability for AI with functional leaders in IT, digital, or innovation. This can result in a “hammer in search of a nail” outcome, or technologies being rolled out without compelling use cases. The orientation should be the opposite: business led and value focused. This business-led approach follows successful adoption approaches in other digital waves such as mobile, social, and analytics.

McKinsey graphics on AI:

2016’s Most Popular 

My most-read blog posts of 2016 were: 

1. Why Anthony Hilton is wrong about DB pensions 

This post, responding to the misguided (in my view) viewpoints of London Evening Standard journalist Anthony Hilton in September garnered by far the most views of any of my blogs this year (around 1400 views). 

An extended version of this article was also featured in Professional Pensions, and also became their most viewed opinion piece of the year

The article must have stuck a chord with readers in the pensions world. We do of course live in pretty challenging times for DB pension funds, with several strong macro-economic headwinds making it harder to deliver the benefits that have been promised. This year saw a vigorous debate around what should, or should not be done to the DB pensions system. This debate was further catalysed by the high-profile cases of BHS and British steel, and the debate looks set to run on into 2017. 

Given the importance of the DB system to the retirement prospects of millions of members I believe a solid debate on some of these important issues is to be welcomed, and look forward to continuing the debate productively in 2017.

2. No Ordinary Collision – the Future of Asset Management

Powerful forces of change are at play in many industries, and asset management is certainly one of them. Technological and demographic shifts will shape the future of the asset management industry, in this piece (April 2016) I discussed some of  the intersecting forces, drawing on a wide body of existing research on the future of work and finance. 

Here are my six key takeaways: 

3. Consulting firms reply to the Work & Pensions Select committee 

In the wake of the BHS pensions story, the W&PSC issued a green paper calling for views on the future of the DB pensions system in the U.K. Given the prominence of this debate and the considerable air-time it’s received this year the responses from the main actuarial & investment consulting firms were considered and insightful. What was also interesting was the diversity of views. Will most schemes pay the benefits promised? Should the role of TPR change? Should there by wholesale change to the system? Will small changes be effective? Is consolidation feasible? These were all questions on which the consulting firms gave insightful, but often differing answers. 
I hope you’ve enjoyed my blog posts this year. I look forward to sharing more in 2017. Sign up to receive updates on new posts.

Roboadvisor Europe 2016 – The Future?

The future of asset and wealth management?

A thoroughly excellent event was organised by Level39 in London on 25 May 2015, featuring speakers from all the key players and analysts in the fairly nascent European roboadvisor scene and around 250 delegates.

My five top takeaways are below, or you can read my storify story here.

1. Mind the 2016 inflection point

Rohit Krishnan of Mckinsey made this point, but it was echoed by others. The growth rates of the two longest stablished US roboadvisors (Betterment and Wealthfront) have stalled somewhat, possibly co-inciding with the robo launches of two large incumbents Vangard and Charles Schwab. With significantly lower AUM than is probably needed to justify costs and valuations, 2016 could be an inflection point, which way will things tip?

2. Customer acquisition cost is key

It’s a closely guarded secret when it comes to individal firms, but surveys and other data in the public domain suggest that customer acquistion costs can be in the region of $300 or higher, but lifetime value of an average client may only be $250. If that’s true, then it would seem to pose a challenge to the business model.

3. But Europe is a bit different

Several of the European robos made the point that the European market is a bit different to the US. With less competition on fees in the traditional advised space, European robos charge in the region of 40-70bps rather than 20bps in the US. This means the breakeven point in AUM might be somewhat lower, 0.5-1bn was suggested.

4. It’s all about the api

There was a fascinating panel covering the tech aspects. The main takeaway being that we have entered a new era of openness, and what’s important is opennes with regard to architecture & api , to enable other components “plug in” to create an ecosystem.

5. Scale & brand are hard

These two comments stood out to me from all the points made by the startups. Firstly, building the right scale to reach 1m+ customers is difficult (more difficult than we thought said Shaun Port of Nutmeg). Secondly, several panellists commented that building the wider brand and customer awareness was key, no-one had really done it yet, and many firms were in a race to try and do so.

that’s it! plenty more I could say (and check out the storify for more).

Taming Pension Risk? 

This article first appeared in the May 2016 edition of The Actuary magazine (here)

The past 10 years have been a uniquely challenging, volatile and uncertain period for UK defined benefit pension schemes. How well have they weathered the storm? We found out by analysing data published by the Pension Protection Fund (PPF) in its annual Pensions Universe Risk Profile or ‘Purple Book’, plus other industry-wide data from leading sources.
The asset side of the story, viewed in isolation, seems positive. Total assets have grown from £770bn in 2006 to £1,254bn in 2015, an average increase of around 5% per year, albeit a very significant proportion, possibly 25-45% of the increase, has come from contributions. Over a period that includes several crises, the asset value has had a relatively low volatility (or standard deviation) of about 5% per year.
Total liabilities have grown even faster though, at an average rate of around 7% per year. The increase is from £790bn in 2006 to £1,516bn in 2015 on the PPF’s measure (which, for most schemes, caps benefits and pension increases), or from around £1trn to just under £2trn based on full benefits.

Furthermore, the volatility has been about 14% per year, almost three times that of the assets – largely a consequence of plunging yields on British government bonds or ‘gilts’, which are used as a benchmark for valuing the liabilities.
Asset allocation has changed in two main respects. Firstly, there has been a move away from equities – from 61% in 2006 to 33% in 2015 (see Figure 1 below) – in particular, a reduction

in UK equities. A slightly different dataset published by UBS shows allocations to equities of over 80% in the 1990s, so the past decade is part of a continuing trend.
As discussed later, this has reduced the level of risk, but it has also reduced the expected future investment returns in excess of the liabilities, by around one-third over the decade. The second trend is an increase in the amount of liability hedging.

When long dated interest rates fall or inflation rises, the value of liabilities increases relative to the assets. This can be hedged by investing in matching bonds, but also by entering into derivative transactions with similar economic properties to bonds.
From a risk perspective, the key statistic is the hedge ratio, which estimates the proportion of liabilities that are effectively hedged (or immunised) against movements in interest rates or inflation. Using data published by KPMG in its annual Liability Driven Investment (LDI) Survey, we estimate that the average hedge ratio has increased over the past 10 years from 15% to around 33%.
At an aggregate level, there have been small second-order changes in asset allocation relating to a greater use of hedge funds and other assets, albeit a more pronounced feature in some individual schemes, and fluctuating levels of cash. Holdings in property of around 5% in aggregate are a fairly constant feature.

Risk change – a conflicting story

To gain insight into the de-risking effect of these asset allocation changes, we ran the data through our pension risk model. Figure 2 (below) measures risk in relative terms, showing the possible change in funding ratio of assets to liabilities at a given confidence level, often called funding ratio at risk (FRaR) and measured in percentage terms, which is useful to trustees in quantifying how benefit coverage could be affected in an adverse period.
In 2006/07, pension schemes were running an average FRaR of 20%, meaning they had a

one in 20 probability of suffering a funding ratio fall of 20% or more over one year.
As expected, it can be seen that the asset allocation changes implied by the aggregate data (a reduction in equities and an increase in liability hedging) have been risk-reducing, with the FRaR having decreased to about 10% by 2015.
These risk levels may seem large, but, to put them into perspective, over the decade funding levels have twice fallen by more than 20% in a one-year period (according to the PPF’s data) – once in the year to April 2009 at the lows of the financial crisis, and once in the year to May 2012 as bond yields fell.
Figure 3 (below) measures risk in another way, in absolute monetary terms, showing the possible change in funding deficit or surplus at a given confidence level, often referred to as value at risk or VaR.

This is often helpful to corporate sponsors in indicating how much the pension scheme could affect their balance sheets.
In 2006, the aggregate VaR was around £300bn, meaning they had a one in 20 probability of suffering a combined increase in funding deficit of £300bn or more over one year (this is based on full benefits rather than the lower PPF measure).
While the VaR is influenced by similar factors to the FRaR, it is also affected by the size of the pension schemes. The overall increase in size of the assets and liabilities overwhelms the de-risking described earlier, resulting in an increase in VaR to £450bn in 2015. Again, putting these figures into perspective, there was one annual period (the year to May 2012) where the combined deficit increased by more than £300bn.

Efficient risk reduction

Will the current asset allocation deliver the returns that pension schemes need to be fully funded? We cannot tell from the aggregate data, as each individual pension scheme has its own specific circumstances, such as funding and investment strategy. We can say that, on average, while funding ratio risks are lower today than in 2006, so are the expected future investment returns.

Many pension schemes want to continue to reduce risk. After all, the risk to the funding ratio described earlier, of 10% over one year, is still substantial in the eyes of the majority of schemes and corporate sponsors. However, they need to generate the same, or increased, investment returns to close their deficits over the next 15 or 20 years.

The solution is to stop viewing investments in terms of fixed buckets of ‘growth’ or ‘matching’ assets and embrace modern-day liability management techniques. This includes the use of derivatives to reduce liability-related risks without needing to invest the majority of the portfolio in bonds, thereby enabling the scheme to invest the remainder in a range of return-seeking assets.

Using the principles of diversification, risk control and downside protection, schemes can generate returns more efficiently – that is, for less risk – than has been achieved previously.

Dan Mikulskis is managing director and co-head of asset liability management and investment strategy at Redington Ltd

– See more at:



No Ordinary Collision

Your 10-Minute Guide to the Future of Asset Management.

Download the full paper here >> no-ordinary-collision-v6-singlepage-HR

It’s always easy to ignore or dismiss forecasts of the way the future may change our industry. Some might seem too obvious, some too far-fetched. We all exist in a daily whirlwind addressing the challenges of volatile markets and demanding clients.

It’s great to take a step back though and take a moment to think about the future of the fund management industry. that’s what i had some fun doing when I sat down to write: No ordinary collision, your 10 minute guide to the future of asset management. I took the time to read through a lot of the great thought pieces out there relating to these future themes (from the likes of PWC, EY, McKinsey and Forbes) summaraise what I saw as the key themes running through them and add some thoughts of my own.
the way I see it, the asset management world looks set to be affected by future trends and themes from at least two sources,
1. the future of work
2. the future of pensions & asset management
Each of these taken on a standalone basis are facing considerable change, taken together it creates some powerful intersections that could really change the way our industry functions and the importantly the key value chains within it.
I’ve spent some time thinking abouut what I think are the key intersections (these are shown and discussed in a little more detail below). you can read the fll report here.
Clearly there’s no single blueprint for success in such a changing and disrupted world. I’ve tried to lay out what I believe are 6 key things asset managers can address. I’d love to hear your thoughts, do tweet me to start a conversation.