Log in


Date: 2015-06-11 21:23
Subject: Merton lectures
Security: Public
The retirement finance lecture that I heard yesterday had a very good statement of what should have been obvious.  He'd already established the many reasons the defined benefit (DB) plans are impractical and problematic.  What he pointed out is that the regulation, goals, and reporting of defined contribution (DC) plans is not appropriate.  DC plans use financial instruments from wealth management backgrounds, and have been reported that way.  This is misleading to customers, regulators, and DC managers.

The goals of a pension plan are to provide a steady income over the period from retirement to death.  DB plans are described in terms of expected income ($/month typically) and percent funding (a measure of risk to income).  This is proper alignment between goal and reporting.  DC plans are described in terms of current value and value growth.  This difference leads to a variety of mis-allocations.  The example he chose was to determine what a "safest" or "lowest risk" investment, according to the metrics. For a DC, the "lowest risk" investment would be T-bills re-invested monthly.  Value would be extremely steady and safe.  From the perspective of anticipated income, this is terrible.  The expected income from such an investment has fluctuated by a factor of five over the past decade.  That's a huge risk to expected income.  A proper low risk investment would be a laddered holding of inflation indexed long Federal bonds, maturing over the course of the retirement.  That provides a steady income.  The wealth value will fluctuate like crazy, especially for younger customers.

This difference is recognized by finance managers and regulators, and they are making changes.  But, the pace of this change is glacial.  I've  only recently seen modest improvements in how my DC plans are reported.  By contrast, my DB plans are perfectly clear about the small fixed pension I can expect, and what the risks are to that steady income.  The definition of "safe" automatic investments remains distorted by using the wealth metrics rather than income metrics.  DC plans still include a "cash" alternative for "safety".  A real "safe" investment would be something like laddered TIPS, but that's not one of the regulated required safe alternatives.

As Merton put it to his Sloan school students: "This is changing very slowly.  You are some of the people who will be doing the hard work of making these changes."
Post A Comment | Share | Link

Date: 2015-02-03 16:44
Subject: Books Read in January, 2015
Security: Public

  • Indonesia, etc., Elizabeth Pisani, Well written kind of travelogue.  Instead of being written by a tourist, it's written by a news reporter/ HIV researcher.  So the perspective and locations are very different than a typical travelogue.  It gives a very nice feel of Indonesia.

  • Faceless Killers, Henning Mankell.  This is one of a the large Wallander series of murder mysteries set in southern Sweden.  I read it to get a feel for how much the BBC Wallander series matches the books that it is derived from.  I prefer the TV series.  These are good but not gripping murder mysteries.

  • No Sooner Met, Seanan McGuire.  This is a long short story, but since it's published separately I'll list it with books.  A nice little story from the October Daye series.

  • Poor Economics, Abhijit Banerjee and Esther Duflo.  This is a kind of "round two" on using evidence and analysis in assessing efforts to deal with global poverty.  Rather than being pure speculations, it takes the results of early RCTs and micro-economic analysis to discuss what is working and what is not working.

  • Collected works of Arvin Alchian.  These are premier quality discussions of various aspects of micro-economics.  It's the kind of stuff that I didn't get in school, and oddly didn't really come across until recently.  It's also two volumes of 800+ pages each. Examples:

    • What is the meaning of Profit, Cost, and Inflation?  I learned definitions by accountants and accounting standards.  These papers look at the underlying economic relationships behind the accounting definitions. It's an interesting and different perspective.

    • how does knowledge or lack of knowledge affect trade?  For example, incorporating non-uniform knowledge distribution and a cost of obtaining knowledge leads directly to predicting sticky wages and unemployment.  Specialists, agents, etc. change the equilibrium price relations.  This work is complementary to the related work of Coase on knowledge and organizational structures.  It has direct relevance in current Internet situations where the cost of obtaining knowledge is changing dramatically.

  • Monsters of Education Technology, Audrey Waters.  She's a "Cassandra of Education Technology".  This is a collection of speeches, rather than written essays.  They were good speeches, but as a result the whole lacks rigor and substance.  Arguments are posed, but not defended.  It's a problem that is inherent to speeches.

Post A Comment | Share | Link

Date: 2015-01-19 19:50
Subject: Books read - December, 2014
Security: Public

  • The New Democrats and the Return to Power.  Al From.  This is a good memoir/history of the revival of the Democratic Party from 1980 to 2000.  Al From was a major insider and while there is the usual self contratulatory tone it's also an engaging historical story.  The publication timing feels related to the Clinton candidacy.

  • Everything is distributed.  O'Reilly freebie.  Just a series of puff pieces about potentially interesting software.  Not all that interesting to me.

  • Peer Participation and Software.  Another freebie. Really aimed at problems of government organizations and how they manage software.

  • Engineering Systems, de Weck, Roos, and Magee.  Intro to system engineering.  It's a nice reasonably complete introduction.

  • The Fourth Revolution: The Global Race to Reinvent the State, John Micklethwait & Adrian Wooldridge. Another view of history and current events from the classic liberal perspective.  It begins with a discussion of prior "revolutions" from Hobbes, Mills, and the Webbs.  Then it argues that the success of Europe from 1500 to present is due to conscious evaluation and change of government structure and purpose.  Hobbes gave the nation a philosophical basis.  Mills added human rights and anti-corruption.  The Webbs took that and created the welfare state.  Then it discusses the semi-revolution of Friedman with his "free to choose" devolution of decision making power.  They forsee next move being towards greater accountability, perhaps creating the epistemic state (validation through success).

Post A Comment | Share | Link

Date: 2014-11-07 14:58
Subject: October Books Read
Security: Public
During the month of October I finished:This turns out to be about a couple software project postmortems.  It's not nearly as useful or informative as I expected.
An interesting ethnographic study of two high schools, one in the Bronx and one in Amsterdam.  Mr Brown spent a couple years teaching in each one.  These are each arguably the worst schools in their areas.  Brown examines the social structures, lives of students, etc.  The similarities are extremely strong despite all the cultural differences.  Probably the strongest factor in failure for students is the pervasive fear and danger.  Their lives are spend dealing with threats.  The strongest indicator for escaping the dead end lives that so many face is having some protective adult continuity.  A strong parent or strong relative is the usual source.  This is not something that can be offered by a teacher.  It needs to be a multi-year experience of having a healthy protective adult that the students need.  It needs to have the continuity of a healthy person to person relationship.SF.  Disappointing.Interesting analysis of situations which were an anarchy, and how order develops.  What were the major factors.  What were the results.  He argues that the claim "anarchy can't work" is easily disproved.  He does not argue the claim that civil government is inferior to anarchy.  He shows what anarchy does accomplish.

The major factors in successful anarchies are the patience of the participants and the extent to which they must maintain long term relationships.Actually about how the IT systems and organizations were structured and tested for the Obama campaign.  The contrast between this level of care to organization, goals, testing, and reliability and the complete failure of similar efforts in Obamacare launch are tremendous.I expected better given the high quality of his lectures and newspaper articles.  The information content is very good.  The book style and organization was mediocre.  Still worth reading.  History and commentary on the whole economic situation since 2005.
Post A Comment | Share | Link

Date: 2014-10-20 18:33
Subject: On design and standards
Security: Public
There is a lot in common between product architectures and standards.  I was pondering that today after someone suggested using Agile methods to create standards.  There are Agile tools that would help in some places, but this sounded more a fad following suggestion than a realistic suggestion.  Actually I think some of the techniques specifically invented for distributed team software development would be of more direct benefit.  I just don't think people are prepared for what agile would really mean.

My comparison of the Agile prinicples with  DICOM's SOP class orientation and it's supplements is:

  1. Customer satisfaction by rapid delivery of useful software

  2. Welcome changing requirements, even late in development (not really, workitems are postponed until requirements look stable)

  3. Working supplements and software is delivered frequently (weeks 1-2 years rather than months many years)

  4. Close, daily monthly cooperation between business people and developers

  5. Projects are built around motivated individuals, who should be trusted

  6. Face-to-face conversation is the best form of communication (co-location) (only partially.  WG-06 is always colocated for the public comment, ballot, and final text preparation.  Tcons just don't work for that.  But a lot of the rest is tcons and email)

  7. Working supplements and software is the principal measure of progress

  8. Sustainable development, able to maintain a constant pace

  9. Continuous attention to technical excellence and good design

  10. Simplicity—the art of maximizing the amount of work not done—is essential

  11. Self-organizing teams (only partially, but not a top-down structure either)

  12. Regular adaptation to changing circumstances

One of the problems with applying Agile to standards is the differing definition of "useful" and "working".  Is a standard "useful and working" if it changes every month?  If not, then what does define "useful and working"?  In comparing it with DICOM and their supplements I decided "useful and working" means that a published supplement can be used as the basis for developing software and that subsequent additional supplements or changes will not invalidate that software (or at least not significantly).

Defining workitems and supplements to match that definition of "useful and working" would be a huge change, and one that I don't think a lot of people would be comfortable with.  Individual workitems and supplements would need to be much smaller in scope, each one being small enough that we get it right (or close to right) the first time.  DICOM is republished 5 times per year, so this becomes an assertion that if you develop against the published standard, you have a very low risk of subsequent publications invalidating your work.

Reminds me of the multi-year battles within DICOM with the RESTful and Radiation Therapy folks.  Both went through years of non-progress (stonewalling?) by WG-06 until they finally broke their work down into small pieces.  Now RESTful is mostly complete and Radiation Therapy is making progress.
Post A Comment | Share | Link

Date: 2014-08-18 10:05
Subject: Some programming
Security: Public
I finally took a break to do some programming for fun.  I put together some Python programs to fetch weather observations (METAR) and forecasts, extract the parts I wanted, and put them into the background window dashboard of "conky".  Conky is similar to the Windows "rainmeter" tool.  It can put various useful status stuff into a semi-transparent window in the background wallpaper.  The other excuse was preparing to help my nephew who is going to be taking an intro course on programming, and Python is the chosen first language.

Now that I've used it a bit I agree it makes a good first language.  All the important basic concepts are there, it's got a huge library of useful parts, it's widely available, and it's a simple syntax.  For example, there is a large library of meteorology tooling.  I could get an open source METAR decoder, so getting the current observation consisted of fetching the METAR (a few lines to do HTTP GET), invoking the METAR decoder to get a complex object, then selecting and formatting the observation fields that I wanted displayed.

End result looks like this:

Post A Comment | Share | Link

Date: 2014-07-27 18:14
Subject: Computers vs Travel this week
Security: Public
It  was a bad week for computers.  During the F2F I was doing battle with a difficult Lotus Notes.  It was in the mood to hang, vanish into never never land, and become as slow as refrigerated molasses.  Repeated kills and reboots were needed.  I don't know why it had problems.  Lotus Notes sucks big time.  I never did get it to work from the hotel, which is odd.  Usually that same hotel has better speeds than RSNA.  For regular web browsing the hotel worked OK.

When I got home the Internet was out.  Come morning I saw Verizon out working down the street.  Apparently a truck tore up the FIOS cable bundle on Friday.  The Verizon crews were there until late Saturday night.  They had to clear the mangled cable, put in a new cable, and then splice the cables.  They said it was 144 fibers to be spliced in that cable.
Post A Comment | Share | Link

Date: 2014-07-18 21:35
Subject: Mis-use of laws
Security: Public
More details need to be disclosed, but there are growing indications that privacy laws like HIPAA are being converted into walls against whistle blowers.  The latest is growing evidence that VA employees who disclose unsafe practices are being threatened under HIPAA.  You can't disclose facts to back up claims of unsafe practices without disclosing at least minimal patient information.  The story is that the VA is claiming that informing your lawyer is a HIPAA violation.

The continual cold war between open source, Microsoft, and Apple hit me again today.  Microsoft will no longer format a USB device that is larger than 4GB with a FAT-32 file system.  In practice you can manipulate block sizes to get much larger devices supported, and I've gotten such devices.  Now it must either be exFAT or NTFS.  The problem with these is that they are not publicly documented and are covered by some US-only patents.

Why is this a problem?  I want to carry around movies, music, podcasts, etc. on a portable hard drive.  I want to plug this hard drive into my Windows, Linux, and Mac systems with read/write access.  That's precisely what Apple and Microsoft don't want to have happen.  Apple will not support any Microsoft disk formats (other than FAT) and Microsoft will not support any Apple formats.  They both want to own their customers.

Fortunately, Linux is a fringe issue, has lots of non-US developers who can ignore US-only patents safely, and has engineers able to reverse engineer a file system.  You can now get a reasonably robust and reliable NTFS read/write support on Linux.  This rescues my 500GB portable that had to be re-formatted.  It's now NTFS.  The Mac can't use it, but I can use network access when it needs files.  That's not great.  But at least I can stuff a couple hundred GB of movies and podcasts onto the little disk.  I don't have to decide now what I'll want when travelling.  I can take everything and decide then what I want.
Post A Comment | Share | Link

Date: 2014-07-16 11:08
Subject: Book series: Brilliance by Marcus Sakey
Security: Public
The first two books of this trilogy are published, with the last coming soon.  I found it from Scalzi's "Big Ideas".  The basic concept is that since 1980 one in a thousand children has been "brilliant" has an extreme savant skill of some sort.  They are otherwise normal, not disturbed like the "idiot savant".  This leads to discrimination, fear, conflict, etc.  They are simple plotting, simple characterization, with a fast pace.  Without the SF context these could be action detective stories.  They're a good easy read.  I'm not surprised that the first one is already under consideration for a movie.  This will make good action movies.  I hope they go for the intelligent detective/action style.

The "big idea" posting is the authors perspective.
Post A Comment | Share | Link

Date: 2014-07-02 09:29
Subject: Future of mobile health
Security: Public
This came across my web feed while I was starting again on reading MPQ.  I give you the future of mobile health:
1 Comment | Post A Comment | Share | Link

my journal
June 2015