management, private credit, national resources, esg, et cetera. there's quite a volume of new initiatives. this is the third and final presentation that we have for risk management. the last two have been done in the prior two board meetings. those two were first regarding as owe allocation and rationale for our allocation and what we expect in terms of our return profile. we expect to outperform in down markets. we expect to outperform in ordinary markets. we expect to outperform in good markets. we do expect to underperform in great bull markets, although our absolute level of return would be very, very good. the design of the asset allocation to do two things, retrieve high returns and minimize the impact by negative

markets. the second risk management presentation that we had last month was a very complete review of our liquidity profile and how we're thinking of them in terms of the pacing. we have 7.5 billion dollars of current unfunded commitments to the private markets. so we've walked through biasset class and in the aggregate, what that liquidity profile would look like and how it changes over time and that it's our goal that in about 8 or 10 years, is that the distributions, the net distributions, net of capital calls from the private markets will be more than sufficient to pay plan benefits, which would be a great outcome. today we have the third and final presentation. then we will do this annually. today is about portfolio

analytics and also about the exposures by asset class and funds. this is ground breaking material. all three have been new. we also have with us susan and david who are partners in the cofounders of our risk management platform. i'm going to ask anna to walk through the items and susan and david are available for questions. >> thank you. commissioners, bill reviewed that we already covered two large pillars of the risk management framework. we reviewed the strategic at owe allocation or policy portfolios, ran through numbers and scenarios for the strategic allocations, reviewed the liquidity management. today we are going to talk about the risk exposures for the total fund in each asset class.

with that, i'll ask our partners to join us because this work of labor for the last 8 months since i joined, and certain much longer for her. i would like to start to acknowledge every single person in the investment staff who is very diligent about risk management. risk management is not a band of a single person. every person who is underwriting and you see that in our discussion, is very aware of the risk return tradeout. what we will talk about today is, as we put together little by little this investment recommendation, how does it all look as a total portfolio? so we will review three parts.

the first presentation we're covering, performance analytics, we'll do the performance contribution and risk adjusted returns for the fund. our partners from an apc present kindly and regularly, looking at expanded analytics. then we will look at the exposures through geographic regions, countries, and sectors. so when i started, i needed to pull together the data and who do we work with? what does the plan look like? i know there's a lot on this page, but look at the middle of it. in the middle of the page is our asset allocations, strategic asset allocation or the policy portfolio. you could see the equity allocation targeted 31% of the

plan. fixed income is treasuries and liquid credit is 9% of the plan, et cetera. this is what we're solving for which is as we mentioned, this it one of the largest decisions that drives the expected returns. how are we implementing and partnering with? what analytics are we using to support our decisions in terms of evaluating the risk return tradeoff and bringing the proposals to you every month? so, for example, on the left, you will see that in our private investments, private credit relapse -- private equity, we collaborate extensively with consultants. you hear every month cambridge associates. we do a lot of analysis and collaborate with torrey cove and

our custodian provides further analytics through burgess. that's just one piece of analytics. for example, with our absolute return portfolio, we collaborate with blackstone asset managers. we also work with alborne to look at their due dill against and further engage with hfri to understand what goes into the alternative space. so these are just two examples. the other example that i would like to bring up is on the right. we recently instituted as part of our overall risk management framework deep dive reviews for each asset allocation. for public equities, we produced

over 300-page report using five different analytics. style analytics. we collaborated with napc, we put together reports, e-investments. we pulled it together to get the full perspective. the reason we need that many collaborations, because each asset class is unique. what style analytics provides for public equity portfolio, the fact to exposures is very different from what we see when we work with torrey cove and even further difference with kaisa. what we engaged over a year ago is to look at the overall aggregate portfolio analysis. now that we have seen the pieces in the allocation, this is the part. how does this look like?

kaisa has been amazing in this journey to pool the data, to weave all these different threads into the oval quilt. i would like to turn it over to susan to introduce kaisa and talk about the example of how we put it together, the processes, the data review, and also thank the amazing responsive partners in this undertaking, and can i would like to acknowledge susan and david, our cofounders and ceo and president. kana is our representative. he's been answering my calls and e-mails over multiple weekends. megan is sitting there, i call her data queen. she knows every data point for multiple portfolios.

i thank them for their partnership. >> thank you. nice to meet all of you. i was here a year and a half ago. my name is susan. i'm the president and cofounder joined by dave who is also a cofounder and ceo. for those of you unfamiliar, the platform is designed to aggregate exposures across multi-asset class portfolios, inclusive of hedge funds, private he can wet funds, direct investments, coinvestments, and typically complex portfolios. our entire client base is institutional allocators, like public pension plans, endowments and foundations. our relationship with sfers january in january of 2017. that's really when the on boarding commenced. they were in the process of moving from northern truck. so our team successfully migrated all of the historical

data back to the 1980s and uploaded that in the platform. the true on boarding and ramp up began with anna's addition in 2018 where her team has been working diligently to make sure that the data quality and integrity is sound in the platform. we've been working with other sfers service providers, including torrey cove on the private equity side and blackstone on the hedge fund side and melon to aggregate all of this information so that we're now at the point where sfers can successfully report their total portfolio exposures across different asset classes, countries, sectors, currencies. they're able to report total portfolio performance and contribution, and where we're going is in the future, we would like for sfers to be able to rely on us to do better modeling, to set pacing across

privates, and to understand how the portfolio would change if they were to hypothetically add or remove managers, like this conversation has entailed today. so it's very nice to meet you and we're here generally to support the presentation that anna and team will be making. thank you. >> thank you, susan. >> i would like to move on. we have a lot to cover. we'll start on page 5. the reports, unless i indicate, are direct screen shots from the platform. this is the evolution over the last 20 years of assets under management, and you will see the gross in asset management, but you also see the difference in the allocations. we will see when we look at the return attributions that most of the volatility was coming from allocation to global equities,

public equities, and public fixed income and we are now building more diversified portfolio. page 6 shows the same allocation but by weight rather than market value. this is the screenshot. very easy for us to track by different frequency and time line. page 7 shows similar data but zooming it into quarterly data and over five years so we can actually see the deployment rate to different asset classes. moving on, now that we have the data on one platform, we would like to see -- we've discussed a lot about the diversification. what is the contribution from each asset class to returns and when diversification worked and when it didn't work. you see, during the tech bubble

in 2000, 2001, 2002, public equities and private equity portfolios lost a lot in value. however, the fixed income part and real asset provided diversification during that time. and posted positive returns. during the 2008 crisis, you will see that there was no diversification benefit. we will examine it further and understand what does that systemic stress look like for the portfolio. more recently last year, public equities didn't perform well. however, we provide the real assets and public equity provided a lot of diversification. so we can really open up the contribution by each asset

class. >> can you say that last part again? >> if you look on page 8, the last bar chart that says 12-31-2018, this is the 2018 performance, which was flat. public equities were down more than 6 to 7%. however, the total fund, as you will see, was slightly up. that's because of the real assets in public equity provided diversification. >> real assets and public equity saved us. >> significantly. >> to put numbers to that, the s&p was down four last year and global equities were down more than nine. okay. and fixed income was, you know, flat. so private equity last year, our

private equity portfolio was up 21%. our natural resources program was up 23. real assets, meaning real estate plus natural resources was up, i think, 15. so we had really monster contributions from particularly real assets and also private equity and also to a degree, real estate. >> thank you so much. >> and that's what causes that, where you see on the far right, the public equity contributed negative 4%. but our total return was slightly in the positive, that little white dot. we're one of the very few plans -- they were three plans that made money last year, and we were the top ranked in the country because of private equity and natural resources and good selection throughout the portfolio. >> thank you. >> thank you, bill.

we can examine it further. page 10 talks about the performance analysis for 2018. we see 12 data points, data we pulled from custodian, and it's provided by the system. you will see that the first column on the bottom chart table talks about sfers total return in 2018 was a quarter pe percen. we compared to benchmark of 70-30. how did we construct that? we took 70% of global equity net return and 30% of u.s. bloomberg

barclay's average. that performed minus 6.48%. the total performance in 2018 of the funds was 6.73%. we also see, again, only 12 data points here. so some of the statistics aren't that significant, but we could see that downsize volatility was lower for the fund. we also could see, from this analysis, the column left from the right was higher for sfers as compared to the 70-30 portfolio. we did a similar analysis for five years. here, again, using as a benchmark, a simple 70% global equity an and 30% barclay's, we

seek it was almost 18 -- we can see it was 19%. it was 3% annual. we also did that with less risk. we define risk as volatility of the portfolio. also with less down side risk and more positive months. moving on to ten-year analysis. sfers delivered adjusted returns, 70-30 benchmark. remember, the reason we picked 70-30 is because it's close to the risk that we take on our strategic asset allocation. the expected volatility is close to that benchmark volatility or risk. so again, the superior risk

adjusted returns, less risk, a third less risk, and 1.89% additional annualized return over the last ten years. >> if i could commented on that. that 1.89%, when you compound that, you see to the left of that, there's an aggregate outperformance of 40% over ten years. >> and over 30 years, it's even more dramatic. it's 449% compounded. 1.78 annualized with a third of risk. so this is due to many factors that we examined. asset allocation being one of them and teams provide each

asset class provides is another big contributor to this superior return. we move on, unless there are questions, to the second part of the presentation where we'll look at the total fund exposures. we start with regional exposures. you will see that we're underweight europe and overweight asia emerging. the interesting -- we'll dig into that in more detail because, for example, it doesn't tell the full story. when we open up with further capabilities that we work with and we look at the overweight and underweights in asia development -- let's start with that -- in the middle of the page, you'll see japan and hong kong.

both countries are part of asia developed region, but we're overweight hong kong and underway japan within this region. i would like to comment briefly on the benchmark that we chose because it's relevant but at the same time, we just need to -- for some of the analysis, we need to be aware of what we're comparing against. if you remember, we decided to run a cash overlay for the cash we manage. the cash overlay was just 68%. so it's a global equity exposure. 32% bond benchmark. here, the bond benchmark is our first benchmark which is two-thirds treasury and 1/3 barclay. so these are -- the bond

portfolio here is conservative, but it's in line with the benchmark. and the 38% is in line with our cash overlay program that we have implemented today. we would like to move on to sector exposures. again, they are coming from the bond portfolio, and that's the allocation. so let's look at the larger sector exposures. it's information technology where we have 5.4% tilt overweight in information technology sector. we also have an overweight in health care. the real estate and energy are coming from specific allocations

to this sector. next, we worked closely with kaisa to customize how we would like to look at the exposures and what makes them for sfers, for our risk management. we're continuing our exposure in asia, across all classes. what does it mean? how big is it actually? we've been approving multiple asset managers, but across all of the portfolios, how much do we have, say, in china? well, let's define china. we think -- we would like to define it as greater china and include china, hong kong, taiwan, and singapore. with this designation, we see that 3.7% of the total fund

coming from global equities or 897 million. 944 million with in private equity. so over 2 billion of assets are in greater china. mostly across equities, but real assets and private credits are contributing as well. >> about 8.3% of plan assets. >> the other exposure analysis that we can run across total fund and asset class is through net exposure. we work very closely with blackstone asset managers to understand risk exposures on net and gross basis. we understand that some numbers

are reported and not netted. we engage with blackstone and managers to understand what goes into it and monitor it. this is the most important part. we also sort that data so we can aggregate our total fund net and gross exposure. so that's on the absolute return. for all other classes, asset classes, the data comes either from custodians or torrey cove. you'll see the other mismatch, the other gross exposure coming from managers using futures and recent extension strategies. we move on to now just doing similar analysis for each asset class. we start with public equity. public equity is our benchmark.

so we use the exposure versus public equity portfolio. we begin to see the underweight in europe, stronger overweight, more than 5% in asia emerging markets and underweight in asia developed. let's open up by country. again, we see 5.3% overweight to china. significant underweight in japan, germany, and france. slight overweight in india and hong kong. on the sector basis, we have strong yields toward health care in our public equity portfolio, 5% versus the benchmark.

we have strong tilt in the information technology, which is 2% of the weight. underweight in financials and strong underweight in energy. moving on, our treasury performance, this is what i joke. this is why i want no surprises. you will see all the numbers match very closely to benchmark. it's the same duration. yield is the same and the other is lower. we hold -- our treasury portfolio, no credit risk. it's 1 to 10 year treasuries. that's all that we have there. liquid credit, this is where we do take credit risk, and we would like to estimate how much we take that. again, this is where kaisa has been extremely helpful to pool

data from our managers and see what we actually have as an aggregate fixed income portfolio. i am going to move on to -- let me see. yes, on page 23, lower chart, bottom chart, the bar that says average credit qualities. you could see that the -- we are comparing to barclay's ad, which is averaged credit quality of aa, and now current portfolio, it's bbb. that's by design. if we combine it with treasury, it still will be -- which is all aaa, it will be at the barclay's level. the sector exposure, again, we have an allocation to corporate

high yield, bank loans, and bonds which often classify as emerging mortgage bonds. >> can i make a quick comment on page 23? >> several. >> so on the bottom chart here, there are desirable characteristics, and a tradeoff. we have lower duration. we have a significantly higher yield. we have significantly higher coupon and we trade that in exchange for lower credit quality. we do that to rely on the credit underwriting of our existing managers. if the underlying credit is good, we will outperform here. go ahead. >> thank you. that's a very good overview of liquid credit portfolio. we are moving on to private

equity portfolio. an interesting story here. you've been approving one manager at a time. here, on an aggregate, you will see 23% of the private equity portfolio is in asia-pacific as compared to 8% for the cambridge global benchmark. we also have strong overweight in it. 48% is in information technology versus 35 compared to the benchmark. these are -- the more salient yields within our private equity portfolio, and they are by design. private credit, the portfolio we continue to build. it's newer. however, you can see on the right chart it's already fairly well diversified across different sectors as well as

we're diversifying it with geography. we're underweight europe, but we're -- if you actually look at some of the financials and the leases, they include leases and it could be classified. >> can i ask a question on the benchmark? are we talking everything subordinated to senior? >> so this is cambridge's benchmark for private credit. the way they looked at it is, the subordinate -- i will have to talk to how they constructed it, but that's all credit, subordinated credit that's been underwritten. >> okay. can we just -- we don't need an answer now. i know -- >> i can pull up the details what's in cambridge's benchmark. >> that would be great. in our discussion about private

credit, i think we tend to talk about senior credit as opposed to maybe things much further down the capital structure. so if we can just follow up and -- i know you're just using this for informational purposes. >> as a benchmark. >> if we can try to compare credit quality of our asset, senior versus junior, however you want to call it, versus theirs so we have a better understanding that is actually the best benchmark. >> sure. >> thank you. >> private credit. moving on with real estate, it's harder to construct because many managers do not report on the actual assets that they manage. so it's on the total fund. so as a result, the benchmark --

and cambridge warned me that the benchmark data isn't that indicative, but you will see for our portfolio, it's very well diversified across different property types, industrial offices, apartments, retail, and hotels. that's, again, by design. moving on to natural resources portfolio, the ben. benchmark data, it's the same. we wanted to provide some ben. benchmark, but cambridge can only aggregate at the fund manager level. but you will see that across value chain, we have much less upstream than the benchmark. we have more diversified sources

exploring the natural resources. >> when you say upstream, you're talking mining, drilling. >> mining or oil. >> okay. so the extraction of the resources? so when you say that we have more of a tilt forwards -- >> much less. so if you look at the -- apologies. i misspoke. we have strong underweight of 33% versus the benchmark in the upstream. so the benchmark is 74%. this is the chart on the right. first portfolio is 41%. >> the message here for the board is that one of the things we voted on was trying to implement an esg policy and be more conscientious about where that money is going and by having further upstream projects. >> by design. >> that we have materially less

fossil fuel exposure than the index. yeah. >> thank you. >> that is the message of the chart on the right. this concludes the exposure analysis, unless there's -- we're happy to take any questions. we'll move on to the second part of the presentation. >> questions from the board,. >> i'll ask a couple questions. i want to understand, this is all assets? >> can it' it's all assets. >> how do you count the more private equity? by flag or -- >> it comes from the service provider torrey cove where we receive portfolio company information and all of the attributes of the portfolio companies. if, for example, you had lyft as a private company, we would also be pulling the attributes, what

current, what currency, what sector,. >> so it's going that way. great. >> yep. >> i'll save observations for later. thank you. >> any other comments or questions from the board? my only request is next time we talk about this, because this is great. thank you for building this out. i know we will talk about this again. since we're moving into private credit as a strategy to build that out, can we just try to see if we can break out senior versus junior, however you want to categorize it. >> sure. >> we do have that information. >> thank you. >> we'll bring it out. >> could you put on the screen pages 30 and the glossary through 35 and just give them a

couple seconds on the screen so our viewers at some time can kind of get an idea what's going on? >> thank you for bringing it up. again, this is where we -- >> you went a little too fast. >> to make sure that we understand how we define risk and return, how we compound that, what we -- how we define this and calculate it. downside deviations and ratios, which is, again, the return divided by the downside deviation, gain deviation, which is the upside deviation, semi-deviations, correlations and bar value risks, which is the measure of the tail risk. >> commissioners, would it be helpful if we spent another minute or two on this? >> it would be helpful because

one of the reasons i asked the question is -- and i had gone to the back to ask of gov tv, some of our viewers following us like to see the documents on the screen. one of the questions that was asked of me was, is there a way for us to download them directly from watching the sfgov tv. i have to report that i can't. i've talked to the experts in the back. they won't be able to do that. they'll have to take pictures of the screen. >> they're available on our website. >> oh, going to the website. but i wanted to give them a little bit more because sometimes we bypass boards. this is actually very good because it brought me back many years, but it's something you can actually plug values into. >> how about if we spend like -- not on every one, but on a handful of these, we'll spend 30

seconds. >> right. >> okay. >> so volatility measures, the variability of our return one year to the next. okay. and the reason why that matters is was when you experience a large loss, you can -- you will experience a large decline in your funding status. even though we are long-term investors, we do care about short-term volatility. okay. so the good news of our annual volatility is if we went back to the reports earlier is that our volatility is less than 70-30. also in the reports that have been generated on a quarterly basis, we're consistently rated in 30 to 40% volatility compared to our peers. drop ratio measures the returns relative to the risk-free rate. it's another measurement of risk

adjusted returns. our sharp ratio is in, as reported by napc, has consistently been in the top 6% to 9%, if you'll recall the numbers. i know it's in the top ten. >> can we scroll through these real quick? >> annual volatility captures total volatility. but sometimes you don't care about what the upside volatility is. that's good volatility. what you care about your downside volatility. that's what this is. it captures what your volatility is in down markets. so our ratio is another measurement of risk adjusted returns. our ratio is also consistently been in the top 10%. i'm going to speed forward through a few more of these. let's continue on here. correlation, let's spend a moment on this.

correlation measures the historical likelihood that two assets have moved together in the same direction at the same time. importantly is that -- so lower correlations contribute to higher degrees of diversification. okay. sometimes diversification works. we saw that in anna's example from 2000 to 2002. we had some assets that made very good money. we also saw that last year. we walked through that example. the diversification worked last year. a couple of asset classes earned 20% returns when public equity was down nearly 9. there are occasional periods where they don't happen very often where diversification doesn't work well at all. that's what occurred in 2008.

that's because we had a systemic crisis, the greatest shock to our financial system that we've had since the gfc, and it really took governments around the world saying, i'm good for that. i'll bail that out. i'll underwrite that, print money, to really get us out of gfc. let's move on here. so we care about things like what our worst returns are over a month or over a year. number of positive months, we would prefer to be positive every month. the good news is that we've had more positive months than 70-30. value at risk, i don't know if you would like to comment. >> i'm happy to comment on that. this is assuming the normal distribution, what is the expectation on the strong move, one standing deviation, two

standing deviation with 5% of the tail or a 1% of the tail. that's something that tells us when the markets move something like one out of 20 or one out of 100 expected days or months or years, how are you going to perform? that's the estimate that we have. we can go back to -- >> pages 11 through 13 or 14. >> right. so, for example, over ten years or -- like longer term, 30 years, you will see the value risk, a loss of 1% for first portfolio was 5.4% and for 70-30 was substantially higher than 7.11%. i'm looking at the last column on the bottom table.

>> all these metrics are really good. we have the lower value at risk. we have a lower semi-deviation, meaning what your volatility is in down markets. we have a lower down side deviation. we have a much higher sharp ratio, lower volatility, higher annualized returns and total returns. everything on this page is good. is that helpful? , commissioner? >> it's not only helpful, but i think it's a commentary that i like to compliment the staff that work on risks across every sector, everybody who is working on risk in some form or fashion, but i don't think that a lot of our members or the public

understand the amount of effort that goes into compiling these and monitoring the risks and compiling the statistics and working on the statistics. i just want to say, i thank you for all your work and sincerely appreciate it. >> thank you. >> thank you. any questions from the board? comments? commissioner? >> i know we're going to do more of this in the next section, first is a labeling issue. you used page 8, you called it fixed income. we're saying liquid credit. the more consistent things are, the easier to understand it. two, pages 8 and 9, you put in the private credit. somebody went back and captured that. we've been in private credit for a long time. the question is, why did we stop or slow down? here's the point i'm trying to get at with these wonderful charts, colorful illustrated.

it's kind of a so what? thanks for describing everything. how are we going to use them? you can take in all the data which mr. casciato was trying to thank you for. the hard work is how we use this information, how staff will use this information to reduce risk or reduce he can pos reduce exp. it's not simply about volatility. if we've decided to be different in the market, we have to state that. >> uh-huh. >> do we need more tools to change it? >> do you want me to answer that now? >> maybe you have a lot more to talk about, but let's lead to that point. >> sure. >> and there is more to say. for fixed income, the fixed income here is a combination of treasury and liquid credit. so whenever there is a fixed income, it's 9% allocation, 6% allocation to treasuries and 3%

allocation to liquid credits. everything that's treasury plus what we have as a liquid credit. so it's the aggregate. i did want to open it up a little bit more than the other asset class because of the barbell approach. >> also, what you see on page 8 is these are reflective of historical weights and not policy weights. yeah. >> any other questions from the board on this item? great. seeing none, we'll open it up to public comment. are there any members of the public that would like to address the commission regarding this item? >> it's all about risk and reward. you have a high risk, high cost, low liquidity portfolio and you only expect a 7.4% return.

the risk taken, you should have a minimum risk of 10%. i've always contended that you only need three investments, stocks, bonds, and real estate. if you invested 10% in san francisco real estate, 30% in bonds, 60% in the dow jones, the average retained is 11%. as i pointed out last meeting, the passive investment, stocks, bonds, and real estate reduce the return of 15% for the last ten years. so you can listen to all the charts you want, and the mumbo jumbo. you're only expecting 7.4% return. i think that is completely ridiculous. thank you. >> thank you very much. are there any other members of the public that would like to address the commission?

please step forward. >> yes. you were talking about the natural resources and the fact that we're under because there's been apparently divestment in fossil fuels. is that correct? can you explain more about that, please? >> ms. landry, as a point of order, what happens is people want to get up and pose questions to the board during public comment, they are free to do so. then as we go on, it's possible we might be able to incorporate those into our question-and-answer session. >> okay. it might be later on? >> yeah. was that it? okay. thank you so much for your time. are there any other members of the public that would like to address the commission? seeing none, i will close public

comment. anything else from the board before we move on to the next item, which is a continuation of this. >> yeah. very good. >> this is really the -- >> sorry. you ready for the next item. >> risk review for sfers total plan. stress testing and scenario analysis. >> this is the fourth and last leg of our risk management presentation. we will do this annually. it's a stress test scenario. it's looking at historical and pro forma results of what impact different market returns would have on our portfolio. i'm going to ask anna to walk through everything. >> thank you, bill. so we are going to start on page 3. we'll give you a background of what we are looking at. there's a lot going on on this page. on the left side, you have

market sensitive. sensitivity. the different markets and stress we apply to this market. in the middle of the page, you see s&p 500 last 20%, minus 10%, plus 10%. on the right, you see the corresponding movement to that market stress or market movement. this is the single factor stress. so we pick a particular factor, energy price movement or currency price movement or interest rate price movement or movement in the global equity markets or you have s&p equity markets. and then look at this particular stress and our portfolio reactions to that stress. remember, we just walked you through all the work that we've done with kaisa to get the

holdings, the exposures in one place for the total portfolio. now, to address commissioner driscoll's question, so what, we can now estimate the sen tiffity of the portfolio. that's one of the two moves that we have at our disposal. bill and i literally looked at hundreds of those sensitivities. these are the ones that we picked. i presented bill with 100 pages of the reports. we reviewed almost every market factor that kaisa provides and looked for what are the meaningful factors that we would like to monitor and understand our reaction to. >> and so we did look at hundreds of these. kaisa has loads and loads of factors that we can evaluate proforma results on. we selected these for the following reasons. it's the most volatile asset.

it's energy. it's currency. it's credit. it's equity. so we picked the big items. that's why you see, you know, some big left numbers here is because we're picking the most serious volatile assets. when they don't do well, what our results would be. >> these are the meaningful ones. we're talking about global equity going down 40% or just to put a perspective, barclays high yield going down 20%. if we assume an average of the spread duration is 4, that means this prep widened by 500 basis points. these are very strong tests. we would like to see, what does it mean for our portfolio. the next page has the results of the same analysis in dollar terms. >> can i ask a question about this? are we just looking at

correlations of asset classes, and are we using average over a period of time? how are we coming to this. >> good question. go ahead. >> so we use -- we can configure that as well. the way we use it in kaisa is the daily asset price movements. for the last year. we assume the correlation that's been realized for the last year. >> and do we have to worry abouh this. if we were to use is this correlation offer the last year -- over the last year, when there are dig down turns, do they tend to be the same? >> core layingcorrelations are . that's why we're going to address that later on in this presentation about what happens when correlations are all one. we're going to get to that.

>> okay. >> not only all one, but we also replace is it. that's why we use the page five. they don't assume correlation. they replay what actually happened where you use a surprise crisis. to answer your question, we actually look at the current holdings in our portfolio. so what susan just described, we know what we hold now in terms of the sect eof the sector, geod the exposures that's been played during that market or over the last market. that's when we looked at those factors that we're stressing. >> does there need to be a

sensitivity analysis within this range of sensitivity analysis? does that make sense? >> it does. i think this is something that is always a question how -- what's the arrow? what are the assumed correlations? that's why we go into extremes. right? that's why you have to look at multiple stress tests, and you have to look at the historical scenario without assuming any correlation. you need to look at the current sensitivity. >> thank you. >> there is no single answer. it's -- you have to look at multiple ways and see -- you have some assumptions. for example, we'll work with blackstone who they define the scenario. they define the correlations in the scenario. that's something we would like to hopefully present going forward and saying not just using the most recent correlations but something that we would like to define.

what is the stress scenario? what is the probability of the stress scenario? what are we looking at? that's the next evolution as you mentioned. >> susan or david, did you want to comment on that? >> no. i think you said that very well. the key thing being the historical stress test actually taking to account what the portfolio owns today and how it performed during that time period from peak. so if, for example, you have this in your portfolio today, it's calculating how apple actually performed during the time period of the crisis that you select. so that's removed from correlations. it's a true actual performance of your holding. >> maybe to dig deeper and look at the stress test on page 6, we look at the sensitivity analysis. so this is one factor. we're stressing msgi equities

going 40% down. and i actually like to start in the bottom right chart that says beta stand alone. what do we call beta? this is the sensitivity of single asset class to that factor. so look at the yellow bar under global equity. it tells me that the beta of our global equity portfolio is close to one, a little bit more than one. what does it mean? if you look up to the right, you'll see percent profit and loss, stand alone, you get the yellow bar for global equity. it says the stand alone global equity portfolio can lose -- predicted to lose more than 40%, 41% under this stress test. so the global equity market is down 40%. our public equity portfolio can

lose a little bit more. so as a result, we have a little bit more of a beta than one. cash is interesting because we run the cash overlay. it is saying that the cash will go down at about -- beta is 58% or something. i'm looking at orange. it stands to lose 22%. >> what this is also saying, these two charts on the far right is the two asset classes that contribute the best in stress environments, or three, are fix th fixed income, but the returns are going to be super low, and the other two are absolute return and private credit. they resist down markets the best, those three. >> you will see the negative beta for identificationed income is two-thirds -- it's actually

whatever we have right now in the portfolio, the treasury portfolio and liquid credit portfolio stand to provide diversification under this stress scenario. right? contribute a little bit to the positive. absolute return is also going to fare well under the extreme equity stress test. it will stand to lose -- stand alone about 6%, and on the contribution, i'm looking at the upper left chart, absolute return which is the blue chart, it's about 1% or 2% on the performance contribution in that stress scenario. >> why is cap so high? inflation? >> no. we have a cash overlay. any cash that we have, it