Speaker 0:
Hello and welcome to this liability driven investing masterclass. I'm joined today by Sofia Sabia, head of investments at the Government Employees Pension Fund. Mark Fair Brother, a member of the Act Society of South Africa's investment committee.

Speaker 0:
Chris, head of liability driven investing and structuring at Ashburton Investments, as well as Trevor Abramovitz, head of liability driven investments at the Old Mutual Investment Group.

Speaker 0:
Welcome to you all.

Speaker 0:
Thank you so quite a technical and interesting topic. Liability driven investing. Chris, Perhaps I can kick off with you. What exactly is LD? I?

Speaker 0:
Yeah, sure. Thanks, Chloe. I think it's quite a wide term, but the best definition I can give for it is that it's in. It's designing an investment strategy to specifically meet financial obligations in the future. So looking at the liability profile, studying it and designing it, an investment portfolio that aims to

Speaker 0:
directly meet those obligations in the future with minimal risk of failure. Well, in a nutshell, um, that's quite a yeah, straight to the point. Um, explanation. Trevor, would you agree with Chris? And perhaps you can expand on the definition of LD I

Speaker 1:
I absolutely agree with with Chris and essentially LD. I is an important investment strategy for a wide range of institutional investors because really, it's

Speaker 1:
the return and the liabilities that that matter. So, for example, if the market does 10% and your liabilities do 12% then this can negatively impact the solvency ratio, the relationship between your assets and your liabilities for an institutional investor. If that investor isn't following an LD I strategy if the investors following, for example, a market based strategy or a liability agnostic strategy, that solvency

Speaker 1:
ratio will decrease as a result of that of that movement. So really LV. I is so important because it provides a framework for investors to understand the various risks that it may be exposed to, like interest rate risk, like inflation risk, like longevity, risk and then decide which of those risks it wants to mitigate or manage so that it can generate stable, growing solvency ratios over time.

Speaker 0:
Thank you, Trevor. So

Speaker 0:
yeah, I agree with with Trevor and Chris, I think the only point I'd like to add to that is that what is crucial is, you know, the the the the shift as it were that you're not

Speaker 0:
benchmarking using an objective measure such as your SWS or caps or any of those benchmarks. Because generally, that's where uh, uh, yeah, I'd say the the the the misunderstanding happens. So you are benchmarking or the benchmark that you're, uh, are basing the performance on is the liability profile itself and not a market based benchmark.

Speaker 0:
So it seems to me as though the liabilities are ultimately what drive the asset allocation and an LD I strategy mark, how would this differ between different institutional investors such as DB pension funds, banks and insurers?

Speaker 0:
Yes. Thanks, Chloe. So, um, I think it's important that every one of these, um, different institutions have different, um, liabilities. So, um, LD I in particular, um, works very well when the liabilities are very well defined in terms of, um, the amount and the timing and the currency. So, for example, something like a DB pension fund, which offers, um, uh, pensions, um, for retirees.

Speaker 0:
Um, these would be long dated cash flows, the cash flows that, um, occur typically monthly. We receive monthly pensions but would recur throughout the retiree's lives. Um, so in that, in that case, you know, a, um an LD I Australia would be very effective because the liability can be very well defined. And as a result, those cash flows in particular. Um,

Speaker 0:
I like to think about them as representing sort of a, um, a fixed income kind of contract. So, um, in in essence, it's the bespoke bond that's representative of of the liability and as such, um, we can construct um, an asset portfolio with a tailored mix of bonds that has a very good match to to that of the liability.

Speaker 0:
Other institutions, um, also share similar liability profiles. So, for example, insurers that offer annuities, um, would also have a very similar liability structure. Um, even, um, insurers that offer risk products such as life insurance, et cetera. You know, those would be debt claims, for example, that occur in the future. You could also structure, um, uh, a liability drive. An investment strategy around that.

Speaker 0:
Ok, Sofia, I see you nodding your head there.

Speaker 0:
Yeah. No, I was just agreeing that it's, uh, a strategy that's amenable to well-defined, um, liability profile that you can estimate, um, you know, upfront, uh, going into the future,

Speaker 0:
Um, and yeah, uh, generally, it's it's, uh, a strategy used for defined benefit pension funds. But there has been a move to to broaden the applications to to other industries, such as insurance and risk products that Mark has mentioned. So I totally agree.

Speaker 0:
Thank you, Sophie. So So, Chris, I want to get an understanding now with DB. I know in South Africa our defined benefits, um, pension funds are slowly but surely now moving over to DC. And it's a completely different context in the UK, which is quite predominantly DB based. But then how would LD I, um, potentially differ between a DC and a DB plan?

Speaker 0:
Uh, thanks. Chloe, I think, um, Mark and Sifiso touched on it quite well. Um, in order to deploy, uh, implement an LD I strategy you kind of need a well-defined liability to work

Speaker 0:
towards. And I think what's challenging on the DC side is we must always remember, despite the fact that we've changed the nature from DB to DC from a member's perspective, they still have expenses and liabilities that they need to pay for. So we sort of need to adapt how we define the liability profile, but it still needs to be well defined. So some of the high level thoughts could be, you know, replacement ratio inflation adjuster type

Speaker 0:
measure to develop this profile. But in inherently, they still have expenses and cash flow, so they still have a liability profile. It's just on us as an industry to figure out How do we define that? So that we can sort of implement liability driven investment for them and give them the best possible outcome

Speaker 0:
through their retirement experience? Absolutely. So Trevor. Sorry, Sofia. Did you want to add something? Yes. I just wanted to add that essentially, for a defined contribution, Um, scheme.

Speaker 0:
You essentially do still have a liability profile. It's just after retirement, right? So depending on whatever you know, type of annuity that a member would prefer with, you know, if it's a fixed or life annuity with profit or living annuity, you'd have a different liability profile. And what you are doing, essentially, is discounting that liability profile to the current present future time now based on market rates.

Speaker 0:
So essentially, uh, what you are doing is you still have a liability profile. It's just very different in terms of uh uh, how you are, uh, essentially structuring, um, the asset profile to match the those liabilities. So it's a slightly different way in the sense that you are are basically structuring an asset

Speaker 0:
profile that will match liabilities in future. So you're matching, you know, an annuity profile that happens after retirement. Now, um, as opposed to having a DB set up where you projected all the, uh, the the the liability payments. And you have an insurer of reserve value that you basically value, uh, sequentially over time. So it's

Speaker 0:
slightly different, but the same principles can apply. OK, thank you very much, Sofia. So, Trevor, now delving into the different types of LD I strategies, we've spoken about the fact that assets and liabilities or the assets need to be, um, managed in such a way that the liabilities, um, can be covered, uh, once they come due.

Speaker 0:
When it comes to vanilla strategies, how would a portfolio of government bonds, um, be managed to ensure that the liabilities are immunised? Um, an additional yield is earned, um, when interest rates, when interest rates change.

Speaker 1:
So as as Mark mentioned, the, um, cash flows of a liability stream represent or resemble that of a of a bond, and that's that's absolutely right. Uh, the problem is that if you look at the the cash flows of the the bonds in the South African marketplace, they don't exactly match or resemble the cash flows of the institutional investors that we've spoken about today's liability streams. So, for example, if you're looking at an

Speaker 1:
sure is annuity book that's got cash flow payments that stretch for the next 80 years, we don't have, uh, cash flow set from bond in South Africa that stretches for the next 80 years. So we therefore have to use, um, concepts and strategies like you mentioned immunisation,

Speaker 1:
where what we try and do is we try and get the same characteristics in our bond portfolio that the liability has so the very similar characteristics and these characteristics get quite technical. But they would refer to things like the duration of your bond portfolio would need to be the same as the duration of your liabilities. The convexity of your bond portfolio would need to be more than the convexity of your liabilities,

Speaker 1:
and what this does is it allows, for example, if there is the parallel movement in interest rates So if interest rates move across the curve in a similar way by down by 1% for example, and your duration of your liabilities is is 10 and your duration of your your assets or your bond portfolio is also 10. It means that both your assets and your liabilities will increase by by 10%

Speaker 1:
if you've got your convexity of your assets or your bond portfolio more than your your liabilities. Convexity essentially describes the curvature of the relationship between the bond price and the and the bond yield, so that if in my example, interest rates were to fall by 1% you'd have your bond portfolio rising by more than your liabilities. Conversely, if yields were to increase, then you would have your bond portfolio falling by by less.

Speaker 1:
And what this immunisation strategy allows

Speaker 1:
and LD I manager to do is to every day change the bond portfolio so that those characteristics of duration and convexity are in line with the liabilities as markets evolve every single day.

Speaker 1:
And that really is key to making sure that the asset portfolio and the bond portfolio move in the same direction and protect the institutional investor against unrewarded interest rate risk and stress on their balance sheet or stress on their solvency ratio.

Speaker 0:
Thank you very much, Trevor. So, Mark, are there any other considerations then that need to be taken into account when selecting which bonds to, um, match for the liabilities?

Speaker 0:
Um thanks, Chloe. So, yes, I think it's there. There's a couple of considerations. Um, you know, Trevor spoke about duration Convex. I think that's sort of the bread and butter for for D I managers in terms of assessing the positioning of their portfolio. So, um, you know, if you had a particular view, for example, that interest rates were against

Speaker 0:
move in a particular direction either up or down, you could position your portfolio to be either shorter or longer duration that and that of your liabilities. And if that view was to sort of play out, then you you could stand to profit. Obviously, if it if it doesn't play out, you would. You would stand to lose. So it it It's a very useful tool in determining how how sensitive your portfolio is to interest rates and inflation risk, et cetera, versus that of your liabilities.

Speaker 0:
Um, so the different bonds that are available on the market will have different, um, characteristics and by, you know, having different weightings or different exposures or different amounts of exposures in each of those bonds that would allow you to to get your duration, um, to manage your duration. Quite well. Um, duration is, of course not. The only metric duration is quite a simplistic, um, interest rate metric. There are other metrics which you can use, um, such as Rand to point.

Speaker 0:
Um, there's also the notion of looking at, um, duration at different parts of the curve. It gets quite complex, but, um, duration is quite a simplified metric, which which allows us to capture a vast majority of interest rate risks that are actually present in the liabilities.

Speaker 0:
Thank you very much.

Speaker 0:
So, then, Chris, I'd like to understand, then going a little bit further down the risk spectrum, perhaps looking at moderate government bond, um, strategies. And Sofia, Perhaps you can comment here as well. How can government bonds then be combined as well, with repo, um, agreements or repurchase agreements to earn, um, that additional yield?

Speaker 0:
Yeah, sure. Um so repurchase agreements effectively, um, are a yield enhancement mechanism or tool that you can use to yield enhance your portfolio. So Mark touched on basically the first stage of how you generate Alpha, and that would be taking active views in the portfolio relative to the liabilities. There's additional risks you can take, such as credit risk. And that's where you sort of bring in this repo type transaction where you sort of

Speaker 0:
sell your government bonds today with the agreement to buy them back in the future at a fixed price. So you you keep your hedge in place, um, or your economic hedge. But you generate cash from your portfolio, so we call it sweating lazy assets because government is obviously paying risk free in our context, then what you do with that cash is you can go buy a high quality portfolio of credit investments um, which will then add yield enhancement to the to the portfolio. Assuming your cost of

Speaker 0:
transacting in repo is lower than your cost of purchasing credit

Speaker 0:
or or the return you'll earn on your credit, uh, investments, you then generate the smooth pick up over time on your government bond portfolio and to Trevor's point earlier, it sort of it. It maintains your solvency ratio but can improve it over a long time in a very stable way. There's obviously certain risks involved with the strategy. So the first one is, um, the credit risk. In order to take credit risk, you need to have quite a sophisticated credit process.

Speaker 0:
Um, that's the first one and then on the side. You obviously, as we learned after 2008, you don't want to leave some of these transactions un collateralized. Um, so you need to have that capability as an LD I manager to manage your credit risk to your counterparts as well as on your, uh, investment side.

Speaker 0:
Thank you, Chris. Um, Sofia, perhaps you can comment on, um, the use of these moderate government bond strategies and the employment of repurchase agreements for the government's employees. So no, thanks. Thanks, Chloe. So as Chris mentioned So essentially, the idea here is that there's levels of complexity even within within, uh, LD I strategies. And, you know, you'll start from one end where you know, the the risk is more subdued. And as you increase,

Speaker 0:
uh, it escalates. And the idea there is that as, uh, uh uh. The risk escalates. So does the expected return or alpha that you to to to achieve from the strategies.

Speaker 0:
Um, So, as Chris mentioned, uh, you can, uh, after you re report out the government bonds that you're holding, put it into high, uh, high rated corporate debt. Or you can even go further down in terms of your risk spectrum. And the idea there is to extract more, uh, expected return, um, so that you can, you know, add alpha in your in your strategy. I will add, though, that

Speaker 0:
the alpha numbers that you are speaking here are not are not a lot. It's not like, you know, like an equity portfolio. But over time, um, it does add incrementally and you can add to the surplus, um, of of a pension fund gradually over time.

Speaker 0:
OK, so it seems to me as though you know the user repurchase agreements and then, um, getting a credit exposure then might, um, assist in earning that additional yield. But then perhaps, Mark, you can comment on how these LD I strategies then employ leverage And what are the risks? What are the risks associated with this leverage and How can this be managed?

Speaker 0:
Ok, thanks. Yes. So I think, um, whenever you're using, um, repurchase agreements or financial derivatives, Um, there's a potential for you to create a short exposure, which, um, some could be some by some definitions, could be considered. Um uh, leverage. Um, and when you're when you're employing, um, leverage. Um, Chris touched on it a little bit earlier. There's there's a There's quite a number of of risks associated with financial derivatives and repurchase agreements.

Speaker 0:
Um, there is credit risk to the actual counter party to the derivative. Um, which is why you do do need collateralization agreements in place, but the actual, um, having some sort of handle on when collateral requirements might be placed on yourself is actually very, very important because it it poses a liquid potential liquidity strain on the LD I portfolio.

Speaker 0:
In fact, um, that is actually partially some of the the the problems which which gave rise to the UK Guild crisis sort of last year, um was was the speed of of those collateralization requirements from the from the derivative and repurchase um, agreement counter parties.

Speaker 0:
OK, Trevor, perhaps you can comment there with regards to the employment of leverage. I know this is a point that we didn't really want to, um, elaborate too much on, um, but perhaps how we could play out for, um, pension funds such as, uh, a defined benefit pension funds as well as insurers

Speaker 1:
So that

Speaker 1:
use of derivatives such as, uh, futures and and forwards, uh, for example, allows a, um, pension fund or insurer, for example, to be able to gain exposure to the economics of, uh, a bond without actually going in and purchasing that that bond and thereby

Speaker 1:
be able to obtain those, uh, interest rate sensitivities those characteristics that we spoke about previously so that you can align the the the asset portfolio with with the liability portfolio. The the beauty of employing such a strategy is that you don't need to actually commit the capital

Speaker 1:
to purchasing the bond. You can rather commit the capital to these yield enhancing strategies. These growth strategies that we we've spoken about if we look at it from the perspective of the FSC A in in South Africa, uh, we wanna make sure that we're not in a a net short position, and so what we need to be thinking about doing in terms of responsibly employing,

Speaker 1:
um, strategies that include derivatives is limiting the amount of derivative usage limiting the amount of of of leverage and making sure, therefore, that the assets that you're investing in with that capital are able to offset the derivative exposure appropriately and very importantly, be able to prudently manage the collateral associated with those derivative contracts.

Speaker 0:
OK, thank you very much. Uh, Trevor. So then, um Chris, I'd like to understand what would then determine the decision of whether to over under hedge if you have a certain interest rate? Um, outlook. Um, is there a decision that you know, growth managers or even LD I managers would, uh, decide to over under hedge their their liabilities.

Speaker 0:
Look, I think it's very much mandate driven, so every single fund is probably slightly different in their risk tolerance. And it's our responsibility to adhere to that. As as LD I managers, we don't want to sneak risk into their strategies that they're not expecting. But assuming there was a budget for it, a risk budget to take take active views on the LD I portfolio, and it would sort of fall as

Speaker 0:
part of your overall investment global macro view. So you'd have to have a good understanding of the economics to make a directional pull on interest rates globally. Then how that translates into S a And from there, if you've got high conviction in your view and you've got space to do it, then you could then say OK, if rates are coming down, I want to overhead. If I think rates are coming going up, I can under hedge. But I mean, it is it is important to bear in mind Um, LD. I is a risk mitigation strategy.

Speaker 0:
So the active views you'll see in an LD I portfolio are nowhere near as, Let's call it prevalent as you would in an Aldi mandate where the guys are trying to beat the benchmark We we are much more focused on, you know, protecting that solvency ratio and improving it over time, as opposed to trying to shoot the lights out on one call. If I can put it that way.

Speaker 0:
OK,

Speaker 0:
Yeah. So I'd just like to add, uh to that. So essentially, the construct that's followed, uh, particularly by defined benefit pension funds, is that in your entire asset pool that you'll have, you'll segment it into two broad groups. So first is what you call the growth portfolio. Then second, the de risk or hedging portfolio, the extent to which you put you know, the assets into each, uh determines essentially your risk appetite.

Speaker 0:
So a rising portfolio is essentially where your LD I strategy would be, uh, placed in terms of your asset, uh, profile and

Speaker 0:
and where you place your assets in that, you know, rising. Uh, portfolio is where you basically don't want to take unrewarded risk If I was to call it that. So that's where you're not taking risk. Where you're taking active risk is in your growth portfolio. And, you know, the growth portfolio would have more risk, uh, assets with a higher risk profile. That's where you put your equities property, et cetera. You'd put it in that uh, uh, a pool where you take,

Speaker 0:
uh, rewarded risk. And then, um, your your lower, uh risk. Uh uh, pool, in terms of asset profile will go into your de risks and then, in terms of, you know, the sort of consideration that a pension fund would make uh, as to how much you put into each of those pools.

Speaker 0:
This is I mean, that is your growth or your rising pool would be things like your your market rate or is a yield curve. Right. So if you have an elevated yield curve, you know, it may generate opportunities for you to be able to, uh, get into, um, uh, LD I strategies at a cheaper rate. If I would put it that,

Speaker 0:
uh, is your pension increase policy in the pension fund? So, for example, if your pension fund does, uh, target, you know, like 75 80% in, uh, inflation increases,

Speaker 0:
um, it would determine the extent to which they would employ an LD. I strategy. Because what an LD I strategy essentially, uh, helps a pension fund to do is lock in, uh, that that inflation over time? Because, you know, one of the key, uh, tools that are used in LD i strategies in inflation linked bonds. Right. So when you, uh, employ an LD. I strategy essentially, uh uh, use the term simplistically allow you to lock in that

Speaker 0:
inflation over time. Another factor is would be, you know, uh, more you know, funds specific terms such as the ratio between your active and, uh, your pension, Uh, members. Um, and obviously the more pensioners you'd have, the more you'd want to. Uh, uh, uh, allocate towards the de risks in strategy. And, uh, other factors such as your funding level are also quite crucial. Because essentially, if a fund has a, um, a higher funding level,

Speaker 0:
um, that is your assets exceeds your liabilities by a a AAA significant or or material margin. It does give that fund. Um, I'd say a bit more. Uh, um, space from a risk point of view to To to to put more assets in the growth, uh, portfolio versus you,

Speaker 0:
your your your de risking portfolio. So, yeah, so those are just some consideration. Um, from a risk profile point of view in employing LD I strategy. OK, thank you. So So I think I'm going to touch on some of the shorter term risk considerations. Um, that a pension fund might need to take into consideration a little bit later. Um, but Mark now I know, uh, Sofia mentioned, um, a growth portfolio as well as a de risking portfolio.

Speaker 0:
What are the risks of having separate, um, asset management teams for both, um, your growth portfolio as well as your your hedging portfolio.

Speaker 0:
Um, I mean, I think one of the risk would be, um, the growth. You know, the growth manager won't necessarily have, uh, cognizance of, um, liquidity requirements. Um, of the actual liability itself. So, um, what What typically happens, though, is that those liquidity requirements are normally funded from the LD I component or the de risking component that you mentioned. And as a result, um,

Speaker 0:
if the liabilities are actually being met out of the out of the LD I portfolio, you know that proportion relative to the growth portfolio will actually start to shrink over time. And then there often needs to be regulator Rebalances that need to occur to make sure that, um, the hedge or the E, the E the effective, um, the effectiveness of the LD I strategy in terms of, um hedging interest rate risk and the liabilities and managing towards the liabilities is actually maintained.

Speaker 0:
So it's a bit more active in terms of actually shifting money around in in the strategies. Um, the other consideration is the liabilities themselves So, you know, uh, for for if the if the liabilities are particularly difficult to hedge. And what I mean by that is that potentially they've got quite significant duration. Convexity, um, the more

Speaker 0:
underfunded you're going to go related to those liabilities in the rising portfolio. Um, the less effective the actual hedging component will become, and you may start to care team interest rate to interest rate now starting to emerge in your your de risk. So there are a number of considerations. It is a It is a strategy that is, that is adopted, uh, quite frequently in terms of segregating the the the the different strategies. Um,

Speaker 0:
but but yeah, so So, um I, I think with the with the growth component, um,

Speaker 0:
provided that the the growth managers are able to meet certain hurdle rates that are implied by the liabilities. And typically the strategy can be quite successful. OK, thank you very much.

Speaker 0:
So, Trevor, perhaps you can comment there and then what would also impede, um LD I growth managers from having greater discretion with regards to the allocations that they, um that they allocate in in their growth portfolios.

Speaker 1:
So as as Chris mentioned, It really is a mandate dependent. It will be a function of what are my portfolio growth objectives? What is the the risk tolerance of the the client mandate? What is the the client's time horizon?

Speaker 1:
Because investing, for example, in uh, deeply liquid asset classes may very be very good for a client with a long time horizon. Not so good for a client that wants to achieve growth over the short term and may want to be doing something else with their portfolio or pool of assets. So those are the factors that I think need to be considered by the the the growth manager, Um, when thinking about which sectors or asset classes to be, uh, giving the client exposure to

Speaker 0:
OK,

Speaker 0:
so it seems as though the liquidity profile of a of a pension fund is incredibly important. Um, perhaps Chris, maybe we can just touch on it. We don't need to discuss it, but, um, some changes in regulation, um, that have come to the fore that might affect the liquidity profile of pension funds. Um, liability streams.

Speaker 0:
Yeah, sure. Um, so, yeah, the two part system giving people early access does present It's not a problem. It's just a new risk. Um, it's a change, Um, that we have to sort of see how it materialises two marks points. Um, usually we fund liquidity in these strategies out of the LD I portfolio. And if there is a big drain, let's let's say not drain a big call on liquidity or massive change in the experience of liquidity required by the fund through time. It may warrant a re look at, um,

Speaker 0:
how much goes into LD I How much goes into growth because you don't want to make the LD I portfolio too small and then to touch back on Trevor's point. It does have implications for E liquid asset class investing. So harvesting a liquidity premium because you've got time in the old days to sort of say, you know, I'm gonna be in this for 10 years. Um, it becomes a little bit more challenging to make the asset allocation calls now because

Speaker 0:
sort of you won't be able to sell your E liquid investments that you might find yourself four or five years from now in an oversized position to E liquid investments as opposed to liquid investments so I don't have all the answers, unfortunately, but, um, there are These are the sort of the things we're trying to figure out as we go forward to To make sure we don't, uh, don't, um, trip ourselves up in a couple of years time. Thank you, Chris. Trevor, did you want to comment there?

Speaker 1:
Yeah. Just to to add that the LD I strategies have been extensively used by DB pension calls. And it's not crystal clear how the two part system is going to be applied in the the case of Of of DB pensioners as far as DB actives go,

Speaker 1:
the kind of take up by DB active members we expect will be a lot less than on or for a DC member. Uh, pretty much because of the the strength of that DB fund and the benefit that it provides to

Speaker 1:
members. It is certainly, we believe, um, less likely that a a DB active member would want to necessarily give up that benefit even partially by accessing their their capital. So from a a DB perspective, our initial take is that the impact of um, the two part system is likely to be less uh, than for the DC member.

Speaker 0:
So, Mark, we've spoken about the importance of hedging interest rate risk. Um, what is the importance there? And was this one of the causes for, um, the most recent, um, banking crisis? Uh, with regards to the Silicon Valley bank failure?

Speaker 0:
Yeah. Thanks, Chloe. So, um, yes, I think, uh, managing interest rate risk is, you know, a very important consideration. Um, and in particular, it it plays out when, um, one needs to actually liquidate assets. Um, and you have to do so at an opportune time in the market.

Speaker 0:
Um, and you've potentially mis mismatched your your interest rate characteristics of your liabilities and and your assets, Um, and that can cause you to actually crystallise losses and potentially in the long term, not be able to meet all your liabilities.

Speaker 0:
So in in Silicon Valley's, uh, Silicon Valley Banks case, um, they had quite a significant material mismatch between their liabilities and for the bank, That would be, um, deposits that they received. So they they have a liability to obviously, um, pay back those deposits when when the depositors call on them

Speaker 0:
and the assets that they, um, put that funding in was typically long dated Treasuries and mortgage backed securities. So you had a A quite a significant mismatch between very, very long dated, um, US Treasuries, which go up to, you know, 30 years plus and, um, your very short dated liabilities, which would be your your depositors,

Speaker 0:
Um, And, um, you know, in Silicon Valley Bank's case, um, that came about there. There was an issue there where the Fed was, you know, raising, Um uh, policy rates and that had a had a knock on effect into the treasury market, where yields started to to climb quite significantly, meaning that their assets declined in value. Um, far, uh, by by a significant amount, um, putting quite a lot of pressure on their balance sheet

Speaker 0:
now the depositors. And it's difficult to pinpoint one specific reason for SV BS failure. But, you know, there was There was there was some contributing factors that that kind of landed in its in its collapse, and one of them was definitely the the concentration of of depositors or all in the sort of venture capitalist sort of space. Um, tech startups, et cetera. Um, so the the speed of on which those depositors actually call on their money was was quite, um

Speaker 0:
um, alarming, I suppose. Or or quite significance, I've heard it being referred to as a as a bank sprint, as opposed to a bank run. And and quite simply, there was There was just not enough. There was not sufficient assets to actually, um, liquidate at a, um at a at a reasonable price because of the the decline in those treasury, um, yields Sorry in those in those treasury, um, bonds, the value of those treasury bonds and therefore they couldn't actually meet all the,

Speaker 0:
uh, the calls from those depositors and ultimately the bank. The bank folded in in a sort of classic, um, bank run. And I think one of the important things that it also highlights is with liquidity risk in particular, we we often think about, you know, US government bonds. We think about US Treasuries as being highly liquid instruments. But,

Speaker 0:
um, it's important to realise that they do carry market risk. So it it depends on when you need to liquidate those assets. And if you need to liquidate them in stress markets, um, you actually end up with not a lot of excess. Um, you So you end up with quite a lot of excess liquidity risk that that needs to be managed.

Speaker 0:
Thank you very much, Chris. Perhaps you wanted to comment there, and then how did this spill over? Was there a spillover effect? Um, on other banks, um, particularly, uh, with maybe Credit Suisse and the U BS. Um, purchase there.

Speaker 0:
Oh, look, I think I agree with Mark So on SVD, um, there's a lot of stuff that happened that then translate to the a LM mismatch, ultimately taking them out the asset and liability mismatch ultimately causing

Speaker 0:
or being a big driving force behind the failure or the It's called a folding, As you mentioned, um, but a lot of stuff led up to that point. So, I mean, if you look at their share price, not hedging, did really well for them for a long period of time. And I think this is one of the

Speaker 0:
the problems with the risk management is that when you start to relax it, um, unfortunately, it doesn't bite you tomorrow. It takes time. Um, and you get into this false sense of security that maybe it's OK. And then something bad happens. Um, Credit Suisse is, uh I'll give you a slightly different situation. Um, I think we we we talk about ESG a lot when we invest, but, um, there was a lot of red flags around the governance there with fines being paid et cetera, which ultimately eventually translated to a massive problem for them. Um, I wouldn't say it's the same problem,

Speaker 0:
but you know, to Mark's point, even though you're a very big entity and you invest in highly liquid securities, when you need to sell is actually quite a big factor on whether you survive a storm or not based on how you've matched your your balance sheet. And I think, Yeah, there's a lot of lessons to be learned there. Um, but you are not much to add to what Mark? OK, thank you very much, Chris.

Speaker 0:
So, So, So Liquid liquidity seems to be, um, one of the most important considerations for large pension funds, Um, in meeting their their liability payments when they come due.

Speaker 0:
Now, with the GP F being, um, such a large if not the largest pension fund in africa um, the large pension funds are increasingly being expected or looked to, um, to allocate to infrastructure development. Um, here in South Africa, as well as through impact investing, say, in the private market.

Speaker 0:
But the nature of these assets, um, are quite illiquid. How would you navigate, um, regulation and supporting increased allocation to these asset losses?

Speaker 0:
And thanks for that. So, you know, with the regulatory piece, um, in South Africa, there has been, I would say, uh, fairly accommodated regulatory environment with regard to infrastructure, I think that allocation was increased to about, uh, 45% across, you know, all sub asset classes, as it were. So from a regulatory part, that's definitely encouraged.

Speaker 0:
Um, from a liquidity point of view, it's It's it. It can be a bit of a challenge. Um, for a general pension fund, Um, and, uh, particularly because a lot of the the big infrastructure projects at least, uh, or the infrastructure exposures are mainly in the unlisted space. And, um,

Speaker 0:
you know, you get a higher return on that because of the liquidity premium itself.

Speaker 0:
There's obviously, uh, uh, other benefits, uh, which, which I'll touch on shortly. Um, the other challenge is with regard to infrastructure type assets is is as well as the complexity involved. You know, um,

Speaker 0:
being, uh, well appraised of what's happening in the, uh, reap bid windows is is, you know, is not AAA simple task. Um, particularly, you know, for a small, uh, a fund, you know, that doesn't have a large, uh, investment team or, you know, has not outsourced, uh, its investment strategy to an investment consultant. So, tho, those are some of the challenges, uh, related to infrastructure. But however, um, I would say there are some benefits.

Speaker 0:
Um, when you look at, you know, uh, the the the the the Asset Profile, uh, of infrastructure assets over time. So you have the the term, Uh, I won't say duration because, uh, you know, technically, that's a very different thing. So the term of infrastructure projects is is generally very long term in nature.

Speaker 0:
And that does match some, um, of, uh, the the liability profile of of pension funds, particularly with a lot of, uh, active members or, uh, active or open pension funds. Uh uh, as it were as well, because the GP F is an open, uh, DB fund. Um, so you do have, uh,

Speaker 0:
uh, members who are active in getting into the fund now, And you have to be able to provide pensions in, you know, 50 40 years time in future. So, uh, long, uh, long dated assets are are very, um, welcome. And, you know, they're very hard to come by, um, as well.

Speaker 0:
There's the inflation linked nature of, uh, the infrastructure assets where, uh, inflation related increases are sort of in built into the cash flow profiles, and that does allow you to to have real, uh, returns over time.

Speaker 0:
Um, there's also the the the the characteristics of a low cash flow volatility, which is generally welcome. Um, you know, it does give some buffer over time, and it gives, uh, as well, uh uh, there is diversification as well. Um, you have an asset class that,

Speaker 0:
um, whose profile is not linked to listed markets as well. There's obviously as well um, the the so societal benefits of infrastructure where it contributes to long term economic growth and basically provides a bedrock from

Speaker 0:
which the economy, uh, can grow from. So it it is, um, I'd say a favourable asset class, but there are, uh, several nuances. Um, that, you know, uh, would make it difficult or difficult proposition for for, um,

Speaker 0:
smaller pension funds. But there are, I'd say products, uh, which the market has come with, uh, that allow, you know, some sort of aggregation to happen. So even if you are a smaller pension fund, you can allocate, uh, and those assets are pooled. And, um

Speaker 0:
uh, those are are being invested into, uh, infrastructure, uh, assets. Yeah. Thank you very much, Sofia. So then I want to understand this infrastructure allocation. Would this form part then, of the, um, hedging portfolio? Or would it form part of more a, um, a growth strategy.

Speaker 1:
So that's an excellent question. It really does depend on the type and nature of the infrastructure investment itself. So if the infrastructure and asset was paying CP I link cash flows and you had liabilities that were CP I linked in nature. Then, actually, you can have this CPR link infrastructure investment as part of your hedging portfolio.

Speaker 1:
It may be marked off. The inflationary bond curve in a similar way to the liability is being marked off the inflationary bond curve. And so both from a valuation perspective and from a cash flow perspective, you can get excellent hedging benefit from this infrastructure investment, while at the same time as my colleagues have discussed providing excellent societal benefits and excellent growth for the pension fund

Speaker 1:
at the same time. If the infrastructure investment, uh, pays Java or floating rate, then you can use that within a derivative strategy or within a portable alpha strategy to maintain your government bond, hedge in place and then port in that that alpha that spread over Java over the the government floating rate and port that alpha into your LD. I strategy to provide an extra pickup over and above the the government bond curve.

Speaker 1:
What we're seeing now with more, more recent um uh, infrastructure deals is is hybrid instruments where you're converting CPR linked at a certain term into into Jabar cash flows. And again, these are the sort of instruments through the right sort of structuring that can be well incorporated into a a pension fund to achieve the overall growth

Speaker 1:
target. I think one of the the things that you know you've touched on is around the the importance of understanding liquidity within the overall uh, risk framework of a pension fund and the important thing to note, um, different, for example, to the UK pension fund. Example. Where LD I managers were relying on those government bonds or those guilts to provide liquidity.

Speaker 1:
You're not gonna necessarily get the liquidity from South Africa infrastructure investments. And so an LD I manager will take that into account when they understand the liquidity requirements of the fund don't necessarily rely on the infrastructure allocation for liquidity. Rather rely on the remainder of the asset pool

Speaker 1:
to deliver that liquidity and, very importantly, understand how that liquidity is going to react in times of extreme market stress. So even if you are at that 45% allocation to infrastructure, as R 28 allows, make sure that you can meet the liquidity requirements that the fund may have over the foreseeable future.

Speaker 0:
Absolutely. Thank you very much, Trevor. So, Mark, then I want to understand then, from this liquidity perspective, what determines the amount, um of liquidity planning that goes into, um uh determining when and how much. Um, these assets can be allocated, say, for example, to private assets or less liquid um, infrastructure

Speaker 0:
OK? Sure. Thanks. Thanks, Chloe. So Yeah, So, I mean, we've we've talked about interest rate risk. We've talked about inflation risk, but at the end of the day, you know, an LD I strategy, you know, has to actually pay out its its liabilities when they actually fall due. So arguably liquidity risk is probably one of the most important risks to actually manage in a in an LD I strategy, it's non-negotiable that it it can't meet the obligations and when when they actually fall due. So,

Speaker 0:
um, in structuring a liquidity profile, appreciating liquidity risk, the LD I manager would would look at the liabilities first and foremost. When are the actual cash flows due? They would need to look at the instruments that are available to them. Um, what do they contribute to the to the to interest rate hedging? What do they contribute to inflation hedging? What is kind of the expected cash flows that would emerge from those from those instruments?

Speaker 0:
Um, are they employing any derivative overlay strategies with purchase agreements, et cetera, um, and in stressed events, as as Trevor pointed to, what is the kind of behaviour that we could expect

Speaker 0:
from those particular contracts in terms of potential margin calls or collateral calls. Um, and then how are the actual remaining assets in the portfolio expected to behave in those stress environments? Would we Do we expect to be able to be able to liquidate them potentially in those stress environments? What kind of assumptions do we need to make about those stress environments

Speaker 0:
in actually constructing? Um, the actual portfolio. So there's quite a lot of things that go into store of liquidity risk planning. And just to ensure that we can be comfortable that, um, with the amount of assets we have in the portfolio and that would inform then the allocation and that we could actually go deploy into a liquid assets.

Speaker 0:
OK, thank you very much, Mark. So, Chris, I want to bring you in here with regards to some of the regulatory changes. Um, that came, um, came into being, um, post the UK Guild crisis. And what have been some of the changes that have happened there?

Speaker 0:
Oh. OK. Um Thanks, Chloe. So yeah, I guess to understand what went was a bit different in the UK to S a. We need to appreciate that they do, they do a similar strategy to us. They just do it a lot more aggressively and effectively. When we talk about Rio's futures forward, Um, we'll use a certain sizing, say, 2025% somewhere around their subject to risk appetite, whereas they used 80 to 90.

Speaker 0:
Now it's not. I don't wanna say it was reckless because I don't think it was. They sit in a very different world than us. Their interest rates, historically over the last 10 years, have been much lower than ours. So they were trying to find yield. But more importantly, um, they sort of they, um they have a less volatile bond market if I can put it like that. So their market value in stable times is much more stable than ours is. And even in uncertain times, just by virtue of us being I'm an emerging market, our bond market is more volatile.

Speaker 0:
So South African LD I manager will be far more, uh, will feel on a daily basis, far more on the collateral side than let's say, someone who's running in a very stable bond market where they're not constantly posting and receiving collateral because things are sort of, um, calm.

Speaker 0:
I guess the risk was running that much more leverage or more repo transactions or more synthetic exposure than us is that if you do have a market crash like they experienced, you are now forced to liquidate all your asset holdings to cover your collateral causes or break your, um, like we mentioned earlier. You need to buy credits. You have to sell your credits in what is arguably

Speaker 0:
quite an unattractive market, and you end up in quite a lot of turmoil. So a lot of questions I'm guessing for all of us came while this was happening in the UK and would have noticed we none of us in S a actually had this problem. And the simplistic reason for that is we just are not nearly as exposed as they are. Um, I think Trevor mentioned it, Um, probably 10, 15 minutes ago. But

Speaker 0:
sorry. Excuse me. We don't do this as aggressively as the first world does. And it's because of this that we sort of protect it. Um, it's not that we are smarter or better than them. We just live in a very different market and therefore we just don't We would be more conservative on these types of allocations.

Speaker 0:
I think it is important to keep the FSC a as we do our reporting or as pension funds do their reporting aware of how much of this sort of activity is going on in the industry because we also don't want to create systemic risk. That, in essence, is what happened in the UK where the sort of the Reserve Bank had to step in and sort of support.

Speaker 0:
OK, thank you very much.

Speaker 0:
Uh, Mark, if I may, just if I may just add to that. So I think one of the big differences in the UK versus South free care is also just the extent of, um, the funding levels of DB pension funds in the UK versus that in in in South Africa. So, um,

Speaker 0:
in South Africa, we in general in general are are far better funded, um, than than our UK counterparts. Um, so I think you know Chris, Chris is right. I mean, they were they were employing more aggressive strategies, and and part of the reason was was to try to get out of that. Um, that underfunded scenario and sort of,

Speaker 0:
um, take on investment risks to try and improve, um, improve outcomes and improve solvency ratios and taking on that investment risk sort of within the LD I strategy in South Africa. We, as Chris said, we also do the same thing, but to a far lesser extent, it's It's certainly viewed

Speaker 0:
or as a primarily a de risking tool. So LD I is primary a de risking tool, and we will take certain, um, additional risks in the form of adding some growth in the within the, um, LD I strategy. But but not to the same extent, um, as well as done in the UK

Speaker 1:
and maybe just to to chip in from my side to one of the the key things that, um is being thought about, um internationally is the type of collateral that is pledged under these derivative and these these republications. So one of the problems in the in the UK guilt crisis was the fact that collateral needed to be posted. It needed to be cash collateral. So the natural place to turn was the the UK government bonds. The guilt

Speaker 1:
sell the guilts to generate the cash to be able to post as cash collateral. The problem, as we've discussed, is that those guilt prices were falling so dramatically, and the fact that they needed to post as collateral sell the government bonds to post more collateral exacerbated the problem. It was a self reinforcing spiral. If

Speaker 1:
there was an allowance to be able to post bond collateral rather than cash collateral, you wouldn't necessarily have had an exponential selling of guilt. And when you think about it from a regulatory perspective, the form and the type of collateral needs to really be carefully thought through so that you don't have these software enforce enforcing spirals, um, occurring exponentially across the market when markets are unravelling.

Speaker 0:
OK, thank you very much So, Sofia, then would it then be all or nothing? When it comes to a DB plan, your liabilities are 100% matched. Then by your hidden portfolio, do you over hedge and what goes into the decision? Um, as to the strategy to follow.

Speaker 0:
Thanks for that, Chloe. So it's it's not all or nothing, Um, and as mentioned earlier on, um, so it's essentially a thing around risk appetite, right? So there's several considerations that uh, would determine the extent to which you derisk as it were. Um, the first is, uh, your funding level that you have as a as as a pension fund.

Speaker 0:
Um, if you do have, uh, uh, uh, a sizable funding level. Then, you know, maybe there's not you. You can basically take some risk off the table and, you know, then basically segment your portfolio to say, OK, we want to, you know, get, uh, increase the surplus over time.

Speaker 0:
Um, and you know, derisk everything else. Um, but if you you you you you do not have, uh uh uh A good funding level. You know, the the the calculus does does change quite materially, as as Mark mentioned, um, you may be forced to to to employ a more, uh, growth driven strategy to catch up,

Speaker 0:
um, on on on your on your deficit or reduce your deficit. Then there's other factors which you know, I did refer to later, uh, earlier on around, um, you know the age or maturity of your pension fund, uh, particularly the split between your active and and and and and your pension. And so if you do have a lot of actives.

Speaker 0:
Uh, you know, if they are, um, sort of more mature and, uh, towards retirement, then it may be worth your while to put more assets, uh, within your de risks and portfolio so that, you know, um, your liability profile is more matched with your your assets over time.

Speaker 0:
Um, And then there's also the the the the crucial aspect of, uh, the the actual yield curve. Environmental market rates in general. Uh, where if you have, you know, higher yields, uh, at a particular point in time across the curve.

Speaker 0:
Um, you may, uh, be able to to to to take. You know, uh, uh, uh uh, uh. Make use of the opportunity. And, you know, derisk most of your portfolio, Um, at a particular point in time. So those are some of, uh, the the the the considerations as well as the size of the pension fund as well. Right? Because, um, you know, the larger the fund,

Speaker 0:
um, you know, such as the GP F. For example, if you are employing, uh, you know, like an active uh uh uh uh um,

Speaker 0:
you know, a bond management strategy, such as, uh, mentioned, um, from an LD I portfolio. You you there may be limits. You know, um, if you are a large fund, because then if the larger you are, then you know, the longer it takes, uh, for you to to sort of reposition, um, some of your your your your asset allocations. So, yeah, So those are some of the considerations, uh, that you would, uh, take into account. It's not all or nothing.

Speaker 0:
Um, I don't like thinking, uh, of things generally from a binary approach. Um, it's more of, you know, the extent to which you put, uh, into one versus the other

Speaker 0:
Trevor. So then from the perspective of an underfunded, um, pension fund, um, or LD I, um strategy, how would an underfunded, um fund then approach, uh, LD I investing?

Speaker 1:
So maybe the the best way to think about that is through an example. And if you think about a pension fund with liabilities of, say, 100 rand and the duration of those liabilities being five, if you've got a fall in interest rates of, say 1% then those liabilities are gonna increase by 5 to 100 and five.

Speaker 1:
Let's say that we decide to allocate 50 rand to an LD. I strategy, then what we need to do to have the same interest rate characteristics is take the duration of that bond portfolio from 5 to to 10 so that if you've got a a fall of 1% that actually your bond portfolio will also increase by five

Speaker 1:
and thereby change in the same way as the liabilities have changed. And that really is key in terms of being able to change your your duration, change your interest rate sensitivity.

Speaker 1:
Given the liability set and given the allocation that you are required to work with and LD I can therefore work very well in an underfunded mandate.

Speaker 0:
So, Chris, considering then where we are in the economic cycle, um, globally interest rates have peaked. We possibly going to starting a moderation, um, of interest rates. Uh, at the beginning of next year, Possibly. How is this influencing LD? I, um, managers, investment strategies?

Speaker 0:
Yeah, sure. Um look like we mentioned, there's obviously risk budgets in these strategies. But if you had a high conviction that rates would start to come down, um, real rates as well you would start to look to position your book a bit longer if you have space. Um, naturally, you can talk about the credit allocation, how you do your overlay strategies, um, trying to be a bit more conservative as there are some calls from that, Potentially a global recession coming through. Um, some of these things sort of need to be factored into the decision making,

Speaker 0:
I guess from an absolute level of rates, it is a good time. And I've seen this from clients to reallocate to our rebalance more towards your LD i portfolio while rates are high. Um, the logic is just, you know, it's cheaper to buy the hedging portfolio. And now, while rates are high and then you'll benefit from that as rates come down. So it's just purely a cost based benefits, um, sort of argument. But yeah, there's a lot of

Speaker 0:
there's a lot of stuff happening in the world, and I think it's a unique opportunity with, um, real rates of between four and 5% to sort of start locking in some of that hedge. Um, as Sofia said, you don't have to do an all or nothing approach, but it is worth looking at. If there is, you know, sort of concerns on other portfolios to reallocate back. Then reset yourself before the next cycle starts. OK, thank you very much. Trevor. Perhaps you can comment and then as well, just with regards to the allocation, Um, in your credit portfolio.

Speaker 1:
Well, I, I have to agree with Chris. I think that you look at where where current spreads are. Are are trading for, um, your inflation in bonds at 4 to 5% real across the the curve. Um, it really does seem like it. It is a a sensible time to be thinking about either allocating to an LD I strategy or to rebalancing and up waiting your your your LD I allocation

Speaker 1:
Not to say that South African bonds aren't or don't have a higher risk premium attached to them. You know, we are have and do have a lot of political uncertainty. Uh, a a growth path is is is worrying, um, and and and so

Speaker 1:
the attractive proposition that we've spoken about in terms of investing in, um, South African government bonds does need to take that into consideration, but it really does depend on where the investor is on their journey to de risking and achieving their their growth targets. From a credit spread perspective, we've seen very big demand across the fixed income continuum from bond funds from income funds for credit.

Speaker 1:
This has driven down the the credit spread levels, which means that managers have to be very judicious in terms of selecting the right credits to be investing in so that you can outperform your portable alpha targets. You can outperform your LD I targets and ultimately achieve the growth that's needed within an LD I context.

Speaker 0:
OK, thank you very much. Any comments from you Mark with regards to the outlook, uh, and interest rate changes?

Speaker 0:
Yeah, I think I think my other colleagues have said it said it perfectly. I mean the the the current environments. Um, although there's there's plenty of uncertainty, a bound a bound, I think, um, you know, locking into to real yields at sort of 4 to 5% appears quite attractive. At least if we look back historically, Um, so you know, those those funds that have not considered any LD I strategies, um, before should you know, should

Speaker 0:
actually consider this as an opportune time to to have at least some allocation. You know, if they are nervous again, it it doesn't have to be a drastic reweighting of your strategic gas allocation away from your current strategy towards a you know, LD I strategy it can be incremental. And as there's comfort develops with the strategy, you can actually, um, you know, increase the allocation of the time.

Speaker 0:
OK, thank you very much. Mark and then Sifiso, um, what other risks are you now starting to take into consideration as we move into a new phase of the market cycle?

Speaker 0:
So I'd like to reflect as well on on the current environment, because, you know, we've not had such a high inflation environment in a while and you know it. It it It gives you, like, sort of the perfect storm, where, you know, because of high inflation, you have higher, you know, rates on the short end. And because of that, you know, there's earnings pressure. So, you know, your growth portfolio is not performing as it should, but inflation is high. So what's ended up happening is you You have high inflation, right? And if you have a pension fund,

Speaker 0:
uh, and you pay inflation linked, uh, sort of liabilities or pensions. Um, you are not able to afford it because you didn't have growth to support that, right? So it basically heightens the need, Uh, you know, for more liability, cognizant investment strategies going forward,

Speaker 0:
Um, which will allow you, you know, to do that more easier in future. So that's that's basically, uh, my you know, the big thing, Uh, that that that's in my thought. Obviously, rates would moderate over time. And, you know, you'll have, you know, your growth as its starting to to get into the swing again. Um,

Speaker 0:
but, for example, in, in, in, in, in our fund, our average duration is is is more than 15 years, right. So our outlook tends to be a bit more, longer term, Um, and, uh, thinking more about, uh, you know, stressing time points such as this one where you have high inflation, but your growth is is also subdued.

Speaker 0:
Ok, so So I'm gonna ask you one more question just to get an understanding of how you stress test, um, your liabilities and assets in A In a environment such as this, um, what new, uh, components or factors are you? Are you stress testing on?

Speaker 0:
So it's it's it's generally trying to, you know,

Speaker 0:
have more. I'd say diversification, right? So allocating more to infrastructure as it were, um, and not and and also moving away from listed markets as well, Because you've had, um, this situation where there's a dislocation between, you know, listed markets performing, whereas the real economy is not

Speaker 0:
So we looking at, um, or or or or or or or actually applying or implementing target allocations within, You know, enlisted markets. Um, which lend themselves to, you know, uh, growth in the real economy, Uh, which, uh, then leads to growth. Uh uh, economic growth. We've got a unique sense,

Speaker 0:
uh, in particular for our fund because most of our a UM. Is is in the country, right? Um, whereas other pension funds would allocate, Um, I think right now it's it's it's 45% offshore. We very much on the lower end of that. So, you know, the growth of the country itself, uh, is is is more top of mind as well. Uh, in our end,

Speaker 0:
I wanna thank you all for, um, contributing your insights into liability driven investing and demystifying this very complex topic. I really appreciate your time.

Speaker 0:
Thank you. Thank you.