COVID-19 BLOG #2 – Author – Alex Stitt

This week:

• Pandemic Dynamics – how’s it progressing in countries and are containment and suppression steps having any effect?
• A “retrospective” on border control; and a surprising, late, source of infection in Australia.
• An update on economic prognoses.
• What the hell is The Market thinking?
• Best and latest information on the dynamics of the disease itself – to inform your modelling.
• And; A bloody good rant.

DISCLAIMER

A few people I’ve chatted with have the impression I’m doing deep research and have some sort of special knowledge. Sorry, simply not true.

This is all fast, furious and possibly superficial as the situation continues to unfold and develop every day. But it does try to use authoritative sources as much as possible.

AGGRESSIVE SOCIAL DISTANCING

At least one member of my family, The General,  has taken it upon himself to practice “aggressive social distancing”. We’re trying to convince him he’s taking it a bit too literally, but it’s an uphill battle.

In fairness, at 2,230 years old, he is in one of the higher risk categories.

But seriously, social distancing is the best tool we, as individuals and organisations have to:

                                                                    
… while helping to reduce infections and save lives.

CASE IN POINT – VO VECCHIO IN ITALY.

This ABC News Report showcases the little village of Vo Vecchio in the Veneto region of Italy. The important details for us, are that, triggered by one positive and the local lock down, early blanket testing of the 3300 residents showed up a near 3% infection prevalence, mostly asymptomatic, and so the village imposed blanket home quarantine on all infected, and then did a second blanket testing after two weeks, finding a much-reduced 0.41% infection rate.

My take-aways?

• We all knew silent infections are more prevalent than we think, this gives one window on how much – 1 cf 89. Plug that into your models!
• Effective social distancing can make a real difference.

WHAT ARE NATIONS DOING?

In the 2006 paper, PANDEMIC …, I noted that all countries, except those that can close borders effectively, are expected to be eventually affected regardless of the standard of health care. I went on to note: “there is considerable debate as to whether any country can effectively close borders” and that the chances of effective border closure reduces if there is a symptom free infectious period in the epidemic disease. 

Fast forward to 2020, and those speculations have proved largely true. Borders leaked way faster than countries could control them – not helped by slow initial responses from both the source of the infection and nations once it became clear a rogue bug was on the loose. With the result that in early March we have a global-reach pandemic.

However, what appeared unthinkable in 2006 is now happening all over with any number of nations, Australia and New Zealand included, completely closing their borders to any non-resident incoming and imposing 14 day self-isolation quarantine on all resident incoming. 

In Australia, until just a few days ago, most infections were travellers or close contacts of them.  Surprisingly, of the suspected country of infection origin, out of the 157 cases named by authorities, the USA was the largest source in a sudden, late surge.

SIDE RANT FOR THE YOUNG ELVES OUT THERE

What a great chart!

Not What a great chart, as in “Pretty, well presented, blah, blah, blah”.

What a great chart as in “WOW!, I never knew that before!”, And WOW!, I can take that away.

Look and learn, elves:

• Early Feb, we shut down Wuhan travel. Infections shut down.
• Early March, Diamond Princess reached its end state.
• End Feb, Iran went feral, then we shut it down.
• Early March, Italy went feral, and we’d learnt from Iran and shut it down. Fast.
• 5 March. The world is going feral.
• 9 March, USA goes feral so fast there’s no hope.
• 19 March. TINA! Australia closes its borders
  ___

  ---

• 25 Jan, one infecton (with thanks to Lewis Carroll for permission to magic up new words).
• 22 Feb, two infectons: 4 cases from the latest in one hit.
• 28 Feb, three infectons.
• 3 Mar, four infectons.
• 19 Mar, 17 infectons.
  ___

  ---

• Those “guardians” in the background were doing a great job protecting us.
• Up until about 9 March where it started overwhelming them.

PANDEMIC DYNAMICS

Previously, we’ve reported the spread of the outbreak across the globe. As it’s now established in almost all countries, there’s little value in continuing that line of reporting. The following table is provided as a final, just to reinforce how quickly the disease has both spread and become entrenched.

More important now is observing how the outbreaks are progressing in each country and whether the extreme measures being undertaken are having any effect.

At the moment, the answer is a very mixed bag. I’m looking daily at new reported confirmed case rates across 40-odd countries – all the big outbreaks, all the big populations, all the continents.  Many outbreaks are way too small to call patterns yet, and testing regimes are very variable between countries, and possibly over time. But this is the picture, ATM.

Of the larger outbreak countries,

• Italy, Iceland and Luxembourg now have confirmed case population infection rates of 0.1% (nearest 0.05%)
• Switzerland, Spain and Norway have population infection rates of 0.05%.

EPIDEMIC OUTCOMES IN AUSTRALIA

Last week I did a “RAG” transformation (rough as guts) of some unofficial US numbers onto the Australian population – you can see those back in Blog #1 but I’m told it’s worth repeating:

With 7.5% of the US population, and assuming our experience matches this view of best guess US experience, it would mean that, in Australia, once the first wave of the epidemic gets properly underway we might anticipate:

• 7 million infections.
• 360,000 hospitalisations.
• 140,000 requiring intensive care.
• 75,000 requiring ventilation; and
• 36,000 deaths.

Australia has under 100,000 hospital beds.

The critical item in the above “best guess” is the community attack rate. This is something each and every one of us and our organisations can contribute to lowering. Every percent the community attack rate is reduced is less infections, less hospitalisations and less deaths.

This week, the federal government has been presented with modelling of how the epidemic might progress in Australia. Unfortunately, I’ve not been able find the official detail, but sources such as Deputy DMO, Professor Paul Kelly and NSW CHO Doctor Kerry Chant have verbalised that the key parameters are:

• Infections in the range 20% to 60% of population. Wide range!
• Population CFR of 1%.
• Fatalities highly skewed to the over 60 and away from the young.
• Peak infections in late April, early May with an ongoing situation through to August.

If we “do the maths” ourselves, as Prof. Kelly evaded, the deaths are substantially higher than my RAG estimate last week.

One senior official is reported as saying our present control measures are “… trying to get the number of fatalities down from six digits to five”.

Clearly, there are some heavy duty RM implications from this, coupled with the economic implications:

• For life insurers, they need to be preparing for significant extra claims amongst their older policy holders.
• For general insurers, some classes of business such as travel, insurance, funeral insurance and business insurances are going to be subject to both increased claims and increased management loads. See Estelle Pearson’s thoughts here.
• For all financial institutions, financial stress will be possibly the highest we’ve ever seen with consequent impacts on bad debts, policy renewals, new business and the costs of operating.
• And for all organisations and for all individuals we need to be laying in 6-month plans, not 2-week plans. It’s too early to be thinking about recovery plans except in as far as making sure that your survival plans don’t smash your ability to recover.
• Brent Walker has prepared some notes on the effects on the healthcare system here. It seems likely to me there will be significant repercussions for health insurers.

ECONOMIC PROGNOSTICATIONS

Since last week’s blog, both here and overseas, entire industries have been shut down. There will be other industries and sectors which need to ramp up during this six-month long pandemic management period, but it’s very hard for me not to presume significant, maybe even profound economic disruption. Most commentaries I see look to some “rebound” in Q3, but most of those were formulated before the extreme measures many countries have taken this week and before the progression of the pandemic during the week.

In Australia, respected economic commentator BIS Oxford (formerly known as BIS Shrapnel) is reported to be recommending the Government needs to provide payments and commercial guarantees equal to 4% of GDP merely to avoid recession.

MARKETS

One analysis that I can’t share with you, concludes the current market, at circa 5000 for the S&P ASX 200, is about correctly priced for a relatively short global slowdown, per the scenario in the McKinsey paper referenced last week.

By implication, the market is significantly overpriced for the more severe scenario.

In your planning, you need to determine what you believe about the economic situation that’s likely to unfold and plan for outcomes accordingly PLUS scenario test around the central case.

DISEASE DYNAMICS

This site: informationisbeautiful.net/coronavirus infographic datapack/ has some excellent visualisations of the disease dynamics, from which the charts below are taken.

Transmission and mortality: No fresh news on either front, but great context.

The following chart is useful for giving us all some perspective. Of course, COVID-19 has only just begun to do its damage so we have to anticipate that this will be a very major global mortality event.

The following chart is new-to-me data putting the Chinese experience up against the Italian experience, essentially corroborating each other.

We still have, frustratingly, no age specific + co-morbid specific data, but these two charts on co-morbidity add some additional depth. Importantly, note the 0.9% No Existing Conditions aggregate mortality rate.

I don’t this think this is new data, all from China, but it highlights the healthcare challenge where 1 in 5 cases may require hospitalisation and one in four of those will require Intensive care.

One missing piece which I hope we can fill in soon is age-specific illness rates. There is a semi-myth that this is an old-person’s disease. That is not true but I don’t have a solid exhibit for you yet. Plenty of middle aged people are getting the disease and getting very sick. Most of them are not dying.

BCP AND PANDEMIC PLANS – WORKING OR PANICKDEMICING?

We’re at a point in the unfolding of this pandemic where it’s starting to be appropriate to ask the questions:

• Have our plans worked?
• Were they adequate?
• Did we start implementing them at the appropriate time and with appropriate urgency?

Clearly, each organisation and risk manager needs to ask this of themselves and there’s little value in someone like me pontificating from the sidelines.

But, all who know me well will be disappointed if I don’t pontificate and so:

AT THE NATIONAL LEVEL:

To me, it’s perfectly understandable that we’re struggling to be able to test adequately for a brand new disease. But it’s almost incomprehensible that we’re struggling to provide adequate masks, swabs, and other basic PPE and supplies for our GPs and hospital staff. These are “every infectious disease, every pandemic” items and we look like we were caught off-guard.

It’s understandable that hospital beds, ICU beds and respirators are not stockpiled waiting to be rolled out. But, it looks like authorities are scrambling to think through, on the fly, what the solutions might be, rather than having mapped that out in advance.

Time and again, from my perspective, I’m seeing the governments introducing measures on a 2Li-2La basis – Too-Little, Too-Late.

No better case study stands out to me than our PM flagging on Friday 13 March and then promoting on Sunday 15 March a ban on open air gatherings of 500 or more (from Monday so we can all do Da Footy on Sunday) but no ban on closed air gatherings of any size. 

On 12 March figures (ie what was visible in Australia on the morning of 13 March) there were 134k cases globally, non-China cases now 2/3^rd of China, 4971 deaths globally, non-China deaths now 60% of China and 128 countries with outbreaks. The non-China components had tripled (cases) and sextupled (deaths) in the week.

“Let’s wait ‘til Monday, eh?”.

One week later, Tassie and NT have locked the rest of us out and we’re talking of suburb-by-suburb lockdowns where needed. Plus we’re chasing revellers off Bondi Beach!

2Li. 2La.

Of course, 2020 hindsight is a marvellous thing.

AT THE COMMERCIAL LEVEL:

Again, 2020 hindsight is a marvellous thing.

My strong impression is that business is reactive and scrambling. Here in Australia (from my cynical perspective) it’s mostly, “I’m only doing this because the government says I have to. Oh, but we really are deeply concerned for your safety and that of our staff and, by gum, we’re giving everything a really good scrub!”.

Let me give two (hopefully non-contentious) case studies – not to pontificate, but to illustrate that business, the world over, appears to have been caught off-guard a full month, or more, after that city in central China none of us had ever heard of before was the headline item on every news bulletin, every day.

To save you too much pain, I’ve truncated these two cases. The background to both is that we were booked to leave for a five week Norway holiday, including a (shudder) cruise through the fjords, departing this Friday coming.

CASE STUDY 1: After pre-alerting the travel agent here in Australia after Italy exploded on 28 February and again when Israel closed its borders on 10 March, it was me who notified them on 13 March that Norway had closed its borders. They didn’t have anything to add until near COB that day, at which point it was “Cruise cancelled”. Surely it’s a problem when your clients are way ahead of you in your business.

CASE STUDY 2: Cancelling other bookings after my Cruise sank without trace or survivors, I looked at cancelling our flight with <anonymous national airline>. Nothing on their website about “trouble at mill”. Business as usual. Even though their country had already closed its borders. It was 4 days before I got an email acknowledging there was an issue and offering some options.

Just to be clear about both these cases:

• At least, 13 days after “Italy infects the world” and induces Europe-wide panic.
• And after a prior month of “Everything you never wanted to know about Wuhan”.

… neither operator was on top of basic facts or responding appropriately in a timely fashion.

Put more bluntly: Looks to me like “someone” (CEO? CRO? Both?) was asleep at the wheel.

LAST WORDS.

“In God we trust. All others bring data”: W. Edwards Deming.

For those who want to do their own explorations and analyses, the “useful links” in the Actuaries Pandemic Resource Centre has links to a number of valuable data sets, including both excel- and tableau- friendly formats.

It might not be on the actuaries links yet, but this Google sheet from a “The Lancet” project is both a frustratingly rich and patchy data set with (this weekend) about 47k case entries, for those who want to explore deeper. It needs a certain amount of “massaging”.

Aggressively practice social distancing (unlike The General and his aggressive social distancing).

And a very final quote for all we risk managers out there from Dr Michael Ryan. WHO Executive Director, Health Emergencies, based on his experiences with Ebola outbreaks.

“Be fast. Have no regrets. You must be the first mover. The virus will always get you if you don’t move quickly. In emergency response, if you need to be right before you move, you will never win.”

CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.