Interview: Gerardo García

The editor spoke with the head of operations at the Bank of Mexico on March 21, 2022.

As we speak, Russia’s invasion of Ukraine continues. From an economic point of view, do the programme of sanctions against Russia present challenges for your reserve management?

The first and the most important issue about the situation is the pain and suffering of the people involved in the conflict. I really hope that it ends very soon. In terms of how it affects reserve management, I can see three different aspects. The first one is about liquidity in the market. These types of events usually drain liquidity and cause more volatility, which means it is much harder to trade and adjust the portfolio to where you want the portfolio to be. This time around, for example, we had our yearly rebalancing just as the conflict started. It was much harder for us to adjust the portfolio, especially for those off-the-run government securities.

The second aspect is that it is much more challenging to trade tactically. Therefore, if you want to deviate from the benchmark for any particular reason, it is much harder as well. Again, this is due to volatility. The third aspect is more about the medium and longer term. The question is whether you should adjust your strategic asset allocation, particularly if your time horizon is not that long, say a year or so, which is often the case for central banks. The reason is that the outlook for some asset classes may be changing notably as time passes by, not only in the short run, but also in the medium term.

Can you give an example?

Sure. Think, for instance, of inflation-linked bonds. I believe inflation globally is here to stay for longer – we’ve already had a shock because of Covid, and we will have another inflationary shock because of this conflict, and the importance of Russia and Ukraine in commodity production in both energy and agriculture. Therefore, inflation will probably stay higher for longer. This calls for considering whether you should adjust your portfolio for such an environment. And think of currencies, too. Commodity exporting countries versus commodity importing countries: where would you prefer to be?

From your perspective at the Bank of Mexico, how has it affected the stance of your portfolio – you mentioned it coincided with your rebalancing?

Our strategic asset allocation is determined by expected market prices. We position the portfolio according to the implied returns of each asset class and financial instrument because we believe it’s an objective way to do so. What are market prices telling us? They are telling us that interest rates are going up, and so we should adjust our portfolio. What we have been doing is to maintain a very low duration – in fact, we have had that short duration since 2021. This conflict is an inflationary shock after another inflationary shock, and this calls for central banks broadly speaking to adjust monetary policy and withdraw some of the stimulus that they have been providing.

What does that mean in practical terms?

Well, inflation-linked bonds, floating-rate notes and money market investments provide some protection from rising prices and rising interest rates, and we have been adding exposure to these. But remember this is not just about the US but also Europe and the UK, where things are more challenging as there is less liquidity in these instruments compared to fixed-rate nominal bonds. Nonetheless, adding some exposure little by little makes sense.

So not a big departure for you in terms of the portfolio mix?

This year we didn’t make significant changes to the portfolio. Indeed, the effects of the conflict may reinforce some of the pre-existing trends I mentioned: higher inflation and higher interest rates, and central banks normalising monetary policy. Therefore, we didn’t make abrupt changes. We are just adding some exposure here and there.

What impact do you think rising yields might have on the trend towards diversification in reserve management?

The diversification trend that we have seen in recent years will continue. This time around what I believe will happen is that some people will start to focus on those larger countries whose economic cycle and monetary policy are going in the opposite direction. There aren’t many, as it happens. China stands out as what the central bank is doing is providing more stimulus, and the expectation is for the People’s Bank of China to cut interest rates or reduce reserve requirements. That’s the type of diversification that people will start to look at. The main driver for this is the rising interest rate environment, which exposes central banks to significant capital losses.

We, therefore, have to do whatever is possible to diversify the portfolio. In our case, we have been adding exposure to Asian currencies for a few years now, and this is because in our view those economies provide the largest diversification benefits. I should be clear: we have been diversifying our non-US dollar exposure into Asian currencies. However, overall our US dollar exposure is probably a little bit larger than a few years ago, and this is also because we expect the US to be the leader in this normalisation process and because the US Treasury market remains the most liquid and deep market across the world.

Could I ask you just perhaps to take a step back and comment on the impact that the Covid-19 pandemic has had on reserve management at your central bank, both from a portfolio point of view, but also from an operational point of view?

The Covid-19 pandemic generated significant challenges for the central bank from a portfolio point of view, as you mentioned, but also from an operational standpoint. From a portfolio point of view, the largest challenge was market volatility and uncertainty about how economies would evolve, and indeed how the pandemic would evolve. Once again what we tried to do is not to come up with our own expectations or with analysts’ expectations, but with market expectations through what is embedded in market prices.

This framework, which has been very robust over the years, helped us to weather this type of environment as it gave us very objective propositions on how to position the portfolio that we could take to our board. We didn’t change the way that we conduct our strategic asset allocation: we relied on the efficiency of markets to tell us how we should position the portfolio.

From an operational perspective, the impact was certainly deeper since people had to adapt to working from home, which brought communication challenges – mainly in terms of organising projects and activities, but also in operations and infrastructure. One very important issue was cybersecurity when working from home.

What were the lessons?

One of the lessons is that financial markets sometimes expose you to unexpected developments. Covid-19 has inflicted so much human and economic pain, but it was also a forceful shock to financial markets. One thing that is always part of our job is to expect the unexpected and prepare for things that we do not know could happen. In that regard, having a balanced portfolio that can do well in different scenarios is part of what we should be doing.

In terms of how we operate, it was very gratifying to see that an institution that is as conservative as the Bank of Mexico could easily adapt to working from home when mobility restrictions impacted our activities. Going forward, working from home is going to be a tool that will contribute to a better work/life balance for people working at the central bank.

Were you impacted operationally?

Nothing really changed in terms of trading or settlement of our operations. The reason is that in 2009 Mexico experienced swine flu, and this was a wake-up call for us in terms of having to guarantee continuity of operations. In 2009, we started trading outside the central bank’s main office in case we couldn’t work there, and we also started working from home. This experience helped us a lot to adapt to the Covid-19 pandemic. It was harder for us to hold our investment committees, but that was probably the one thing that we couldn’t do as much as we did before.

You co-chaired the group at the Committee on the Global Financial System (CGFS) on capital flows last year. Are there findings from that exercise that you think could be especially relevant this year?

I believe the findings will be relevant for quite some time, because the trends, the drivers, the benefits and the cost of capital flows build through time. Therefore, some of the findings reflect what has happened over the past 10–15 years. They are structural in nature and so will be useful going forward.

Let me highlight three briefly. The first is the growing importance of market-based finance vis-à-vis bank-based finance. This means that more economic agents are getting funding from markets rather than banks. That is important going forward.

The second is the changing “pipes” of capital flows. This is the financial architecture that we are living with right now. New players such as nonbank financial institutions have played a prominent role lately. These large investment funds, asset managers, hedge funds, proprietary trading firms, and algorithmic traders can have systemic implications, particularly in smaller and emerging market economies such as the one I live in. For example, nonbank financial institutions are prone to be procyclical and to exhibit herd behaviour. This is a risk today, particularly for emerging markets, where assets tend to be less liquid, and more so at a time when the most important central banks are going to be withdrawing liquidity from the market.

The third finding concerns the policy toolkit to deal with capital flows. This has broadened significantly, and monetary authorities are clearly better equipped today to deal with capital flow volatility than they were a few years ago. That is not to say that using this toolkit is desirable because the first line of defence to deal with capital flows is having strong macroeconomic fundamentals and also deep capital markets. However, if you need to use these policies from time to time, then it is better to have a broader toolkit.

In what sense?

You have monetary policy, you have exchange rate policy, macro- and microprudential tools, but you also have capital flow management tools, and global financial safety net features, such as hard currency liquidity facilities from major central banks, that have also been very, very useful as of late.

Did the work of that group challenge your thinking on reserve adequacy or did you find it confirmatory, if you like?

Well, the group didn’t touch too much upon reserve sufficiency. We know that reserves are self-insurance mechanisms that are very useful. Different countries have different ways to assess whether their reserves are sufficient or not. In Mexico, for example, we have different measures, and we do not put more weight on one versus the other. We clearly stand above the Assessing Reserve Adequacy (ARA) metric from the International Monetary Fund (IMF), which is most often used, and we also lie in the mid-range on other metrics that incorporate the possibility of having sudden stops or even capital outflows.

One thing that was also touched upon in this working group was the global financial safety net, by which I mean credit lines from the IMF and also swap agreements, regionally and with central banks that issue reserve currencies.

Globally, reserve accumulation provides self-insurance, as I mentioned, but it is in my view a suboptimal policy. Look at the cost of carry that many central banks face, and also the exchange rate risk that clearly dominates many central bank balance sheets. This makes me think that having a very large stock of reserves as self-insurance is probably not in the best interests of the world. In my view, having a global financial safety net such as these IMF flexible credit line programmes, regional arrangements and central bank swap lines, which were key in the Covid-19 pandemic by the way, might be a better alternative globally than just having a large pool of reserves in one country or the other.

A portfolio approach then?

Yes. In the Covid-19 pandemic, the swap lines that were provided by the Federal Reserve and European Central Bank (ECB) were key in alleviating funding pressures in US dollars and euros. I think these types of measures are ones that strengthened the global financial safety net and that prevented a liquidity crisis actually. Flexible credit lines, as you know, were also part of the toolkit not only in Mexico – we have had access to it since 2009 – but also Chile, Peru, and Colombia had access during the Covid-19 pandemic.

That helped them anchor expectations for sources of foreign currency funding even if not drawing upon the credit line. And then you also have new facilities such as the FIMA (Foreign and International Monetary Authorities) repo facility from the Federal Reserve, which is a backup liquidity facility aimed at preventing fire sales of Treasuries. My impression is that Treasury securities can be repoed out almost under any circumstance. But you never know, maybe having access to this facility will prevent fire sales in the future and aid countries in having access to US dollar liquidity. The lesson is that international cooperation is key in addressing a crisis.

You mentioned the FIMA repo facility – I wanted to ask about that because in some ways that feels like the newest element.

Yes, and we should be thankful for what the Federal Reserve is doing here. Bear in mind that the FIMA repo facility is more useful for countries that do not have access to swap lines. A swap line is preferable as with it you can access FX liquidity without having to drain your reserves. FIMA is useful, but it is much more like a backup facility as US Treasuries are by far the most liquid instruments. It would be useful under a very stressful scenario, which of course can happen.

Let me add something in regard to the FIMA repo and these types of facilities that central banks had during the crisis. When you want to provide liquidity to the market, it is one thing to provide liquidity against Treasury securities, but when you provide liquidity not only for the safe asset but also for corporate securities, mortgage-backed securities, asset-backed securities and so on, then you really strengthen liquidity mechanisms in a market. Some countries used these liquidity facilities to support their domestic markets during the Covid-19 pandemic, and they were very important in providing liquidity to asset classes that are not so liquid in stressed periods.

Can you say a bit about your approach to strategic asset allocation, which includes market prices as you’ve mentioned. How did it start?

Sure. In 2012, we faced historically low yields and very narrow spreads in the market. This followed a long-term trend of declining yields. We knew it would be hard for history to repeat itself. We couldn’t see yields going lower. Then, in 2013, the “taper tantrum” happened and it was an even louder wake-up call. Finally, we were also accumulating reserves less rapidly at this point. Oil prices came down significantly in 2014 and Mexico moved from a positive to a negative oil trade balance. As a result of these factors, we shifted the focus of our investment objectives from return generation to capital preservation.

Why was the “taper tantrum” significant?

At the time of the taper tantrum there were a couple of weeks where we suffered significant capital losses. At the central bank, we publish the weekly results of the reserve portfolio – a legacy of the 1994–95 crisis in Mexico – and these losses raised eyebrows at the board. As a result, we made some changes.

In a nutshell, we moved from a classical mean-variance optimisation methodology that used historical data to a more complex methodology using forecasts. However, this was a transition – we went from using historical figures, to internal forecasting, then analyst forecasting, and finally to using the information embedded in market prices to guide our strategic asset allocation. Now, during the transition what happened was that when we went to the board with our internal forecasting, they said: “OK, but why should I believe you? What makes me think you are going to be the one that will have the expected return right going forward?” And then we said, “OK, well, you don’t have to believe us, you can believe this pool of analysts.” Then we did the exercise with the analysts’ expected returns, and the board said: “OK, well, but an analyst usually has this view, and maybe they are not so timely in terms of changing their forecasts.” We said: “OK, if you do not feel that comfortable with our internal forecasting or even with analysts’ expectations, then we will look to the markets.”

How does that work in practice?

You have to start with the assumption that the markets you are trading in are most often efficient. From this, it follows that all available information is embedded in market prices, and so you can use those market prices as a guide to the expected returns of each asset class. That’s how this peering into the future with the help of market prices came about. Since 2017–18, we have been using market prices to guide our strategic asset allocation. It has helped us also to be very objective in regard to what we present to the board, and also in guiding our diversification into new markets and new currencies, as well as resulting in a broader asset base that intends to perform well under different scenarios.

What did it mean in terms of diversification of assets and currencies?

Being focused on capital preservation doesn’t mean that return is unimportant. Nor that return doesn’t matter, or indeed that correlation does not matter. As we began to analyse new markets and different asset classes, we incorporated asset classes that enhanced the yield of the portfolio without incurring too much more risk. First came mortgage-backed securities, and then corporate bonds.

When we looked at currency diversification, one thing was very clear: traditional reserve currencies have high correlations with our numeraire, the US dollar. The correlations of the euro, sterling and Japanese yen are higher than those of the currencies of Korea, Singapore and China. Therefore, we looked at that latter group and added exposure to Chinese and Korean sovereign bonds. We also added more names to our list of potential issuers to invest in, including supranational institutions and regional governments.

In time, we also included Australia and New Zealand sovereign bonds as possibilities. Through this time, we also focused on more tactical aspects. We took advantage of some distortions in the FX market and engaged in synthetic investments in US dollars, which allowed for yield enhancement, for example. If you think of capital preservation and the most important risk factors of the portfolio in that regard, the first is interest rate risk. Therefore, we reduced the duration of the portfolio in 2012 and have had minor variations since then depending on the economic cycle and the resulting market-based expectations for interest rates.

Last, in terms of exchange rate risk and commodity risk, there was also an important change in what we did with gold. Gold, as you know, is negatively correlated with interest rates: if interest rates go up, it’s often the case that the gold price comes down. So, we have partially hedged our gold position using options and other derivatives ever since the taper tantrum event.

One piece of advice for any reserve manager looking at new currencies or new markets is to look to the sovereign yield curve ahead of money markets or time deposits, as some people often do. Why? Because there’s also a negative correlation between the exchange rate risk and the interest rate risk of each country, especially in the longer maturities of the curves. If yields go up in Australia, then you have a loss in your fixed income portfolio. But think of what happens with the currency if interest rates go up in Australia – it is more likely that the currency will appreciate. There’s also this diversification component in the sovereign yield curve and the currency.

So, you said your numeraire is the US dollar, but presumably it is also the Mexican peso from time to time when you do accounts?

Actually, no, that is not the case. The numeraire of our portfolio is always the dollar. It is often the case that emerging markets have a foreign currency as their numeraire. This is precisely the reason that central bank balance sheets are significantly exposed to exchange rate risk, as I was mentioning earlier.

For my country, and I expect for many emerging market countries, the most important risk for the balance sheet is exchange rate risk because of what could happen, in our case, with the dollar appreciating or depreciating against the peso.

As a reserve manager, you don’t think in terms of the peso?

Well, you are obviously well aware that your balance sheet is denominated in the domestic currency, but for reserve management purposes you think in terms of the numeraire, so the short answer is that for portfolio construction I think in terms of dollars, and I have been arguing with many colleagues from emerging markets on this! I have a strong view: emerging markets must have a foreign currency as a numeraire. If you do the opposite, you look to the domestic currency – in my case the peso – then the composition of your reserve portfolio will need to have a positive correlation with the peso.

Think about that for a minute: what would happen if you want to have a positive correlation to the peso? You will need to have assets that might be more volatile, riskier, less liquid. And what could happen with these riskier assets in your reserve portfolio with something like a Covid-19 or the Russia/Ukraine conflict? What will happen is that when you need your assets most, when you need your reserve portfolio to grow in value, then you will have riskier assets and the portfolio will probably have losses and/or less liquidity.

That’s not what you want in an emerging market economy. That is the reason I believe that having an asset liability management framework for international reserves is difficult in many emerging economies.

You mentioned Asian currencies. Have you held them throughout that period since 2017 or 2018, or have you added and moved out of some?

We’ve been moving in and out, but I would say that these are small adjustments, only fine-tuning. The bulk of our exposure in terms of Asia is China. Korea has been there, Singapore has been there. China has been trying to open its markets for central banks, although they have been moving cautiously. That has been the case with us too. However, we don’t even deal directly, but currently invest in China through an asset manager.

As far as I can tell, you don’t tranche your reserves at the Bank of Mexico. Is that right, and have you considered it?

That’s right. We haven’t really considered it in the last few years, to be honest. I understand that some central banks often create a liquidity tranche and also have an investment tranche. In our case, there are different liquidity requirements to make sure that day-to-day needs can be covered and that the overall portfolio maintains a sound liquidity position. As a result, we don’t see the need to implement tranching, and we actually believe that tranching results in a suboptimal allocation as you don’t take advantage of all the diversification benefits and correlations between the different tranches.

However, speaking broadly, not only about Mexico, I believe that there is a case to revisit the tranching question, or at least to revisit the size of tranches. It comes with the growing importance of derivatives in FX intervention. If you are going to intervene but you are going to use derivatives, as opposed to spot markets, then you won’t need your intervention currency immediately.

In Mexico, if we intervene it is most likely the case that we will use non-deliverable forwards. And non-deliverable forwards, settled in pesos, do not require you to use your reserves right away. This might be the case for other countries as well, and think it calls for assessing liquidity requirements or tranches in some central banks. It might allow more flexibility.

One trend we’ve observed among reserve managers in recent years is a move into equities, which was once unthinkable for a central bank. In the Bank of Mexico, is that still prohibited by law?

Well, I’ll answer that, but first let me tell you that, in 2021, we revisited the question of having equities in the reserve portfolio. As you can anticipate, the research showed that adding equity to the investment portfolio delivers an increase in the level of diversification, ie, higher expected return for even a lower expected risk. Therefore, from a purely financial perspective, adding equity makes a lot of sense.

Now, back to the Bank of Mexico. The board has been very conservative in sticking to traditional asset classes. We discussed including equities before 2008, but the crisis happened and made people concerned about potential losses. We tried again a few years ago, but the legal interpretation did not favour the inclusion of the asset class. But things change, and I believe including equities makes a difference, so I´m positive time will come in which we can revisit this issue again.

I guess a crucial factor here, of course, is how you think about risk. Can you say a bit about how you measure risk?

Sure. We use two measures. One is value-at-risk and the other one is conditional value-at-risk. For the strategic asset allocation, we use conditional value-at-risk because we focus on the distributions of expected returns and on the left tail of the distribution. We want to know exactly what could happen if things go wrong. Our optimisation model is trying to reach a higher yield for each expected level of conditional value-at-risk. That is actually the proposition of our model.

We also use value-at-risk, but it’s only to measure deviation against the benchmark, and I think that that has also been very, very useful. The benefits of using both VaR and CVaR over some other risk measures is that we get a better sense of the left tail of the distribution. This goes hand in hand with the objective of capital preservation. If you have this objective, then you care much more about the left side than the other parts of the distribution. The limitation of VaR is that once you go beyond it, you do not know exactly what the expected loss can be. However, if you use conditional VaR (CVaR) then you know exactly what the expected loss is.

I also wanted to ask you about a programme that you launched focused on absolute return with external managers – what was the inspiration behind this, and what’s been the results?

The story of absolute return actually goes back to the equity story. When we were told we couldn’t go ahead with equities, we started thinking about how we could enhance the diversification of the portfolio, and came up with the idea of having an absolute return portfolio. The reason relates to what happens with traditional investment portfolios in central banks. Here you need to have a benchmark, and when you have a benchmark you have exposure to risk factors that you do not always want to have.

Think of the traditional benchmark – of maybe one and a half years of duration, say – you have interest rate risk that maybe you don´t want to bear. And with the benchmark you will also have some type of currency exposure that you may not always want, or have credit exposures to, for example, agencies and MBS, corporates, which you may not always want. Every benchmark has some unlikable exposures depending on market circumstances, and of course you cannot deviate much from your benchmark. However, if you have absolute return, then you give the managers the possibility to have their strategies exactly where they want to have them. This has been particularly helpful in environments like the one that we are living in right now, with interest rates rising.

What has happened is that some managers have been holding negative duration strategies, and they have been profiting from interest rates going up. This is only one of the benefits. This extra flexibility obviously adds diversification to the investment portfolio, and also highlights to us some of the strategies that we should be looking at in our internal management of the portfolio. Overall, it has achieved its objective of diversification, the results have been good and it has been helpful in terms of learning for us. In my view, the absolute return programme can grow in size going forward.

What constraints do you put on the external managers?

Basically, the asset universe is the same as we have internally, which was decided by the board. In reality, the universe is actually broader than for us internally because we have some operational constraints around going into some markets. For example, we cannot go into Norway and Sweden in bonds, but the external managers can. We cannot go into some inflation linked bonds because we do not have the operational capabilities, but these managers can. The universe does not include equities but, apart from that, the external managers can invest in a long list of currencies and asset classes. The limitation is in terms of value-at-risk. We have a monthly value-at-risk limit, and obviously we have a hard stop loss, but apart from that it’s very flexible.

I wanted to conclude by asking about your view on socially responsible investing.

The Bank of Mexico has been committed to ESG investments for quite a while. We started investing in green bonds around six years ago. These were supranationals, but now the scope has widened, and we recently included sustainability or social bonds, as well as sovereign and provincial green bonds. We have been growing our exposure to these investments.

Having said that, though, I think that ESG investments tend to exhibit a premium, which is likely associated with their scarcity. As the market evolves and the differences in liquidity are reduced, we can expect price discrepancies to diminish. This will make it more likely for us to increase our exposure to these types of assets. Apart from the holdings that we have, we also have sustainable and responsible investment practices in place in our corporate bond mandate in which we have negative screening practices, where we exclude some sectors.

That’s it in regard to our investments. A couple of years ago we did have a discussion on whether we wanted to include sustainability and responsible investment as one of our objectives in reserve management. We discussed three alternatives. One was having a responsible investor charter as some central banks do, notably in Europe. The second alternative was including a fourth objective in our reserve management: it would be safety, liquidity, return and sustainability. The third possibility was just endorsing responsible investment in green, social and sustainability bonds.

At the time we considered that the incipient development of SRI could, to a certain extent, imply significant trade-offs with the first three objectives: safety, liquidity and return. We chose to endorse responsible investments in green, social and sustainability bonds. But we had this discussion. I should add that the Bank of Mexico is a founding member of the Network for Greening the Financial System, and recently created a new directorate: Environmental and Social Risk Analysis and Policy.

Thinking broadly, SRI is considerably more developed in advanced economies. However, in my view, the low-hanging fruit in terms of projects that will have an impact – an impact on climate change that is – is found in emerging markets. This is where I think we should be strengthening our efforts. Going forward, the challenge for all of us is how to increase funding for the truly impactful projects, many of which are in low- and middle-income economies.