Japan's Government Pension Investment Fund (GPIF) made headlines this week when it confirmed that it would start offering the fund houses that oversee its active equity and bond portfolios performance-related fees from April 1, allowing fund managers that offered genuine alpha to earn more while cutting the fees of those that underperform. 

In an earlier interview held at AsianInvestor's 7th annual Japan Institutional Investor Forum on March 15, GPIF's chief investment officer, Hiromichi Mizuno, explained why the pension fund was planning to implement this new fee structure. He also explained GPIF's interest in helping grow environmental, social and governance (ESG) principles and how artificial intelligence could help it to better oversee its mandated fund managers.

Details of that discussion follow below. 

Q Can GPIF single handedly help ESG flourish in Japan?

Two and a half years ago when we started our campaign to be a signatory of [the UN Principles for Responsible Investment] and also [to] promote ESG through our investment chain, we faced a lot of challenges. And some of the challenges ... I found the most difficult to break through came from the hardcore investment professionals, [who said] that ESG doesn’t help them to make a better return.

And I keep telling them they need to change their mindset because what GPIF doesn’t want is that you make a return tomorrow at the expense of the future sustainability of the company [being invested into] or the system.

The way I interpret ESG is two key words: sustainability and inclusiveness. For those values to flourish we need to have all the stakeholders of the investment chain. GPIF gets money from the pensioners [via] the companies [they work for] and we trust that money to the asset managers and the asset managers, [who] invest it back into the companies. That’s literally the investment chain we create. And in that investment chain all the stakeholders have to support this idea.

So when we introduced the ESG indices we [wanted it to not] only be the topic among the investment professionals but [for] the corporate executives to talk about the ESG and the general public to talk about ESG. And so we really made a huge promotion with our index vendors. And now after one year what happened is in terms of public awareness of ESG, Japan should be ranked the highest in the world at the moment.

We genuinely believe that GPIF’s investment into those indices promotes system sustainability, which makes our portfolio more sustainable. So that’s another way try to introduce an unconventional way of thinking.

Q Is it catching on among other institutional investors in Japan?

A That’s a question you should ask the audience. While the big asset owners and sizeable peers are becoming more [like] asset managers themselves, about 90% of asset owners don’t have the bandwidth to become asset managers. GPIF is unique as despite our size we will continue to follow the business model of this supermajority of less sizeable asset owners.

I really hope that we are setting a precedent for those small asset owners, [in terms of] how they should rebuild their relationship with asset managers and how they promote these kind of things.

Q You have partners with Sony and Accenture to study the impact of artificial intelligence (AI) on asset management. What are you trying to get out of that?

A I like technology personally and since I became a CIO every time I have an opportunity to meet with a [chief executive] or CIO of an asset manager I want to discuss with them how they are going to adapt AI technology into their business. And I haven’t been very satisfied with the quality of dialogue that I’ve had [to date]. And one day I came to realise it may be because GPIF has been regarded as a very dull and unsophisticated investor, and that’s why they don’t want to talk about it. So that’s why I said ‘okay we are going to change our brand … from being a dull and sleepy investor'.

The second reason is, I’m not sure that AI can replace a human manager. But that’s the asset manager’s job to decide. Because we are going to be a user of your service. If you think you had better replace your team with the AI you should do it to give us the best product.

I also started thinking how in five years’ time our team can do due diligence on fund managers if one manager comes to us in a nice suit from Wall Street and says ‘we have 200 analysts and 50 very experienced portfolio fund managers’, and then other people come in jeans who just graduated from Stamford and have just created an AI manager but their performance using the simulation is 10% over the last 10 years. How can we judge each side? That’s the question I try to address and try to make our team prepare for.

With this Sony and Accenture [partnership], I’ll give you one example of what we are hoping to use it for. The one thing the customer most dislikes the asset manager to do is style drift [in their investment portfolios or mandates].

At the moment we don’t have any particular way to detect when their style is drifting, but we have the big data of 15 years worth of daily trading information from asset managers, which we never used other than [for] preparing our annual report. So we hope AI can look into our database to find out when their trading pattern changes rather than wait for the performance report. Then we can call them and say ‘hey your trading pattern has changed over the last two weeks, what’s happened?’

Q Is this another way to negotiate down fees?

A I’m sure everybody is aware that we are renegotiating the fee structure with all existing asset managers. And everybody talks about GPIF trying to use our weight to reduce our fees. I’m very glad you asked this question, because to make this crystal clear we have zero intention to reduce our asset managers’ fees.

 ... When we started analysing the active managers we always face the argument that the active manager is not necessary. Why don’t we just use the passive managers, which [are] very cheap, and active managers as an industry never beat the market. That’s a compelling argument. But I still believe in the humans’ intelligence and I hope the human manager can beat the market.

Unfortunately, historically speaking, less than 20% of our managers ever delivered [the] target alpha they proposed when they applied for our business, and I think that kind of [poor performance] shouldn’t survive. So the quality of our target alpha after the manager communicates with us has to be more credible.

And the second [reason] is that I have a very strong doubt of the alignment of interest between the asset owner and the asset manager, particularly when we discuss ... their capacity. Everybody agreed that when the asset manager’s portfolio reaches some capacity limit the alpha should diminish but, on the other hand, nobody tells me how to evaluate their capacity.

So we retained some [investment] consultants to give us some analytical tools to analyse what could be the capacity for that particular mandate. And after one year they concluded there is no way to do that. So my response to that is, if there’s no way to do that there is only one way to force the asset managers to be serious about it, so our fee structure will become punitive when they take too much money and alpha starts diminishing.

So this structure is basically trying to create a self-governance mechanism. Because I’m very realistic; they (the fund houses) are smarter, they are better resourced, so we cannot really monitor them, so let them be serious about their offering to us, and that’s the basics of the background of the new fee structure.

Editor: on March 26 GPIF confirmed it would introduce a performance-related fee system for its active equity and bond fund managers from April 1. 

Q What feedback have you received from asset managers about the newly introduced fee structure?

A I think the initial reaction was very interesting. ... I sense they feel exactly what you said. And when I say I have no intention to reduce the fee, their next reaction is usually that you are not even covering our costs. And that’s another thing I try to challenge.

Is there any industry other than the asset management industry and utilities where the service provider thinks that they can transfer the cost to the customer? I don’t know how much [research and development capital] Apple poured into [its] iPhone, or how expensive their parts are, I only decide whether the price is worthwhile to add value. And that’s the way every industry is working, so I keep telling asset managers I think the industry should change their thinking.

The very common reaction to that is ‘we have to hire very expensive analysts, to deliver the returns to GPIF. And I said ‘yes a lot of companies pour a lot of R&D investment, but if their product is not appealing nobody will buy it. That’s as simple as that. So we have to challenge all those traditional mindsets of the asset management industry.

At the end of the day ... think how you discuss with your portfolio companies, how you analyse your portfolio company, and apply that to your own industry.

Click here to read the first part of this interview with GPIF's Mizuno.

This interview has been edited for clarity and length.