With current volatility and market confusion, one can find data points supporting arguments for both deflation and inflation up ahead. My current view: Given helicopter Ben being a student of the Great Depression and not being hampered by a gold standard, coupled with quantitative easing (printing), the probability of deflation is close to zero, with inflation being the big risk.
Ok, so perhaps that is obvious, but less obvious is whether gold is an appropriate inflation hedge today. Regardless of what turns out to be correct, this is a topical context with which to test argumentation, reasoning and decision making, and what better way than to have two heavyweights in the ring.
First up a Wall Street Journal article: (read it here)
Then the iTulip rebuttal: (read it here)
We live in interesting times!
Sunday, July 19, 2009
Thursday, July 2, 2009
Expertitis
Just about every human decision about the future is tainted by a gap — the difference between what we think we know and what we actually know. The more expert we are, the wider the gap is likely to be. — Nassim Nicholas Taleb
In an earlier post in 2007 I stated: "To keep perspective I try remember that professionals built the "unsinkable" Titanic, while amateurs supposedly built the arc." OK, so there was no arc, but it's a useful analogy to keep in mind to be weary of experts.
The "quants" have taken an intellectual hammering during this financial crisis, and rightly so. Many were running models with data set of less than 10 years coupled with flawed assumptions that distributions are always Gaussian!
Here is a classic summary of a series of experiments proving the dangers of expertitis: http://www.strategy-business.com/enews/enewsarticle/ac00001
In an earlier post in 2007 I stated: "To keep perspective I try remember that professionals built the "unsinkable" Titanic, while amateurs supposedly built the arc." OK, so there was no arc, but it's a useful analogy to keep in mind to be weary of experts.
The "quants" have taken an intellectual hammering during this financial crisis, and rightly so. Many were running models with data set of less than 10 years coupled with flawed assumptions that distributions are always Gaussian!
Here is a classic summary of a series of experiments proving the dangers of expertitis: http://www.strategy-business.com/enews/enewsarticle/ac00001
Tuesday, January 13, 2009
A Million Words
If a picture is worth a thousand words, imagine the power of a visual mapping argumentation methodology. The result: cheaper, better and faster outcomes.
While evaluating tools and methodologies, (I will write about Austhink's Rationale in a future post), I came across this visual map from Robert Horn at Stanford on whether to deploy a National Missile Defense system: http://www.stanford.edu/~rhorn/a/policy/NMD/NMD.v7.pdf
Content quality aside, imagine the comparing the effectiveness of this diagram with the traditional essay format, in communicating the arguments, scope and completeness of what exists in a single page.
Q.E.D.
While evaluating tools and methodologies, (I will write about Austhink's Rationale in a future post), I came across this visual map from Robert Horn at Stanford on whether to deploy a National Missile Defense system: http://www.stanford.edu/~rhorn/a/policy/NMD/NMD.v7.pdf
Content quality aside, imagine the comparing the effectiveness of this diagram with the traditional essay format, in communicating the arguments, scope and completeness of what exists in a single page.
Q.E.D.
Thursday, January 8, 2009
Corporate decision making
A recent McKinsey survey titled "How companies make good decisions" concluded with the suggested improvement of: "Put organisational goals ahead of business unit goals, and encourage efforts to build consensus between business units".
In many companies where compensation is primarily business unit centric, expecting execution on the above would probably be a bridge too far.
No real surprises though: Better collaboration, methodologies, financial and risk modeling up front, yield higher probabilities of desired results. Another McKinsey article on corporate decision making can be found here.
In many companies where compensation is primarily business unit centric, expecting execution on the above would probably be a bridge too far.
No real surprises though: Better collaboration, methodologies, financial and risk modeling up front, yield higher probabilities of desired results. Another McKinsey article on corporate decision making can be found here.
Wednesday, January 7, 2009
Monday, January 5, 2009
"You come to school to learn!"
I recall this mantra being drummed into us at primary school. Teacher after teacher, as if they were part of some cult.
A while later, a maverick teacher tried to correct this: "No, you don't come to school to learn; You come to school to learn HOW to learn" he preached. That perspective influenced me.
IBM, as part of their e-learning research, found that about half the learning by students is done from other students, the questions they ask, and the water cooler conversations they have. Yet the education system seems to resist change as this Wharton article suggests: http://knowledge.wharton.upenn.edu/articlepdf/2032.pdf
Decades on, my current view is that an improved purpose is to learn how to think. There is so much raw data out there, generated at an accelerating pace, that learning it becomes impossible. The best we can hope for is to learn generic tool sets and methodologies (themselves open to improvement) in order to develop a robust decision making process, which will allow us to deal with infinite data and changes as the situation requires.
A while later, a maverick teacher tried to correct this: "No, you don't come to school to learn; You come to school to learn HOW to learn" he preached. That perspective influenced me.
IBM, as part of their e-learning research, found that about half the learning by students is done from other students, the questions they ask, and the water cooler conversations they have. Yet the education system seems to resist change as this Wharton article suggests: http://knowledge.wharton.upenn.edu/articlepdf/2032.pdf
Decades on, my current view is that an improved purpose is to learn how to think. There is so much raw data out there, generated at an accelerating pace, that learning it becomes impossible. The best we can hope for is to learn generic tool sets and methodologies (themselves open to improvement) in order to develop a robust decision making process, which will allow us to deal with infinite data and changes as the situation requires.
Subscribe to:
Comments (Atom)