SERVING THE QUANTITATIVE FINANCE COMMUNITY

Paul
Posts: 10430
Joined: July 20th, 2001, 3:28 pm

### Re: How good or bad are the pandemic warning models?

Don’t worry about corona. The one to worry about will come from China but will be transmitted via plastic, and cardboard packaging.

Collector
Topic Author
Posts: 4563
Joined: August 21st, 2001, 12:37 pm

### Re: How good or bad are the pandemic warning models?

katastrofa
Posts: 8955
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

### Re: How good or bad are the pandemic warning models?

Fibonacci says that it begins to flatten (this model doesn't have any sense)

Paul
Posts: 10430
Joined: July 20th, 2001, 3:28 pm

### Re: How good or bad are the pandemic warning models?

According to wikipedia Fibonacci doesn't even know when he died or where. So I wouldn't trust his forecasting abilities.

Paul
Posts: 10430
Joined: July 20th, 2001, 3:28 pm

### Re: How good or bad are the pandemic warning models?

I don't suppose there's anywhere we can bet on these predictions. Would that be bad?

bearish
Posts: 5425
Joined: February 3rd, 2011, 2:19 pm

### Re: How good or bad are the pandemic warning models?

There probably is, and not necessarily a bad thing. The World Bank issued pandemic bonds a couple of years ago, and their payoffs depend on:

The PEF covers six viruses that are most likely to cause a pandemic. These include new Orthomyxoviruses (new influenza pandemic virus A), Coronaviridae (SARS, MERS), Filoviridae (Ebola, Marburg) and other zoonotic diseases (Crimean Congo, Rift Valley, Lassa fever).

PEF financing to eligible countries will be triggered when an outbreak reaches predetermined levels of contagion, including number of deaths; the speed of the spread of the disease; and whether the disease crosses international borders. The determinations for the trigger are made based on publicly available data as reported by the World Health Organization (WHO).

trackstar
Posts: 27281
Joined: August 28th, 2008, 1:53 pm

### Re: How good or bad are the pandemic warning models?

A new name:

Coronavirus disease named Covid-19 - BBC Feb 11

Also, the WHO is in China now, helping to evaluate the situation. In addition to that visit,

"Scientists from around the world are meeting in Geneva to discuss ways to combat the outbreak. Dr Ghebreyesus of the WHO said there was still a realistic chance of containing the disease if enough resources were devoted to the fight."

Graphic from the article:

Cuchulainn
Posts: 61537
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

### Re: How good or bad are the pandemic warning models?

Fibonacci says that it begins to flatten (this model doesn't have any sense)
nCov-Fibonacci-forecast.png
Why not just use linear regression?
http://www.datasimfinancial.com
http://www.datasim.nl

Every Time We Teach a Child Something, We Keep Him from Inventing It Himself
Jean Piaget

trackstar
Posts: 27281
Joined: August 28th, 2008, 1:53 pm

### Re: How good or bad are the pandemic warning models?

Fibonacci says that it begins to flatten (this model doesn't have any sense)
nCov-Fibonacci-forecast.png
Why not just use linear regression?
**
Well, she can defend the choice herself, but did you perhaps miss the Disclaimer on the original post:

"I've been reminded about Fibonacci sequences recently and they got stuck in my head - mixing with the fear of catching the disease."

We can easily imagine how that could happen, after a couple of good craft beers and some calculations on cocktail napkins.  : D

Just like Feynman, or even Peter Carr!

Tell me you have never done something like that - maybe it turned out to be brilliant and maybe it was just your imagination....

trackstar
Posts: 27281
Joined: August 28th, 2008, 1:53 pm

### Re: How good or bad are the pandemic warning models?

Odes to hope (for whatever) and fear (of pandemics!)

Just My Imagination - Rolling Stones

That one is my favorite version, but you might prefer a classic:

Just My Imagination - The Temptations

Cuchulainn
Posts: 61537
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

### Re: How good or bad are the pandemic warning models?

I never call him Fibonacci,.. his name is Leonardo of Pisa.

http://www.datasimfinancial.com
http://www.datasim.nl

Every Time We Teach a Child Something, We Keep Him from Inventing It Himself
Jean Piaget

trackstar
Posts: 27281
Joined: August 28th, 2008, 1:53 pm

### Re: How good or bad are the pandemic warning models?

I never call him Fibonacci,.. his name is Leonardo of Pisa.
Or Leonardo Bigollo Pisano...

A rose by any other name is still... a ~Fibonacci sequence. I like chambered nautiluses (nautili ?) even better.

tw
Posts: 1173
Joined: May 10th, 2002, 3:30 pm

### Re: How good or bad are the pandemic warning models?

Fibonacci says that it begins to flatten (this model doesn't have any sense)
nCov-Fibonacci-forecast.png
Why not just use linear regression?
I don't know about everyone else, but the theoretical question that intrigues me is, if using standard epidemiology assumptions
what would be the aysmptotic populatons of susceptible , infected etc. be without intervention.
I have never really studied these models before but had a look through the wikipedia page on compartmental models
I always thought that the "R" in SIR models was a bit ambiguous as it could mean "recovered" as in the wiki page or
"removed" as Britton,  less optimistically, describes it in "Essential Mathematical Biology" (which has sat on my
bookshelf waiting for such an occasion to motivate me reading it closely).

Either way, the SIR model has an unstable, disease free, equilibrium point but also the stable "epidemic equilibrium"
for which the "R" component has the  $T\rightarrow \infty$ solution
$R_{\infty}=N-S(0)\exp( -R_{0}(R_{\infty}-R(0))/N )$
Starting from S(0) = N,  R(0)=0, and expanding around $R_{\infty}/N=1$ gives to 1st order
$R_{\infty}/N \simeq 1-\frac{1}{e^{R_{0}}-R_{0}}$
What did they reckon R0 is again? 1.3?  4?

trackstar
Posts: 27281
Joined: August 28th, 2008, 1:53 pm

### Re: How good or bad are the pandemic warning models?

But as an immediate practical problem: if the quarantine period really should be greater than 24 days, then this means quite a few people have been out walking around who should still have been in isolation with full precautions.

I'm not alarmist, but this is the type of thing where the known unknowns can cause big problems down the road.

That said, I'm going off the pandemic theme for tonight. I "rowed" a 9K on an erg last night and will be doing a shorter steady state work out tonight.

Veni, Vidi, Remigare, Vici!

- Latin proverb

Translation (for the lazy ones among you - hehe):

"Deprived of wind, take to the oars."

Paul
Posts: 10430
Joined: July 20th, 2001, 3:28 pm

### Re: How good or bad are the pandemic warning models?

In SIR the Rs are both dead and recovered+immune I.e. can’t cause any more infections. In going from R to number dead is the easy bit!

You need SIR plus a transition matrix for travel between airports plus diffusion in space. It’s just a credit risk model!

Corid19 low mortality but long quarantine

Black Death is extremely high mortality and short quarantine (I think)

Not the end of the world yet.

Get Jim Murray’s books! All the details are there. They’re quite big. And you can always throw them at anyone who is coughing near you.