Moving to jeremy-chen.org

I'm moving to http://jeremy-chen.org/. Mostly.

I plan to use that site as a "self-marketing website" of sorts and to manage content in a way that I would otherwise not be able to do on blogger alone.

This blog will stay, ostensibly for more provisional ideas prior to refinement. I'll be gradually moving content (I still like) over to the other website. =)

Saturday, April 30, 2011

On Financial Crises and the Lack of Incentives to Prevent Them

I was watching this documentary titled Inside Job. One line of narration got me:
    Since the 1980s, the rise of the U.S. financial sector has led to a series of increasingly severe financial crises. Each crisis has caused more damage, while the industry has made more and more money.
This got me back to this piece of work I did some time ago on showing that a "fixed proportion portfolio" with a re-balancing trading strategy made more money the greater the volatility in the market was. (Of course the finance literature had proven that long ago, but I wanted to do it myself and in a form I was interested in.)

In this trading strategy, the fundamental observation is that if the market goes up and comes back down to the starting level, or down then back up to the starting level, the trader would have made a strictly positive amount that increases in the size of the swing.

There's no need to look up the proof. I've since lost the little paper scraps. Do this in Excel, re-balance with each price movement. You can verify this in simulation. And if you go further and do a full simulation with percentage transaction costs, you'll find that there is a level of volatility that gets you positive profit most of the time, and that that more volatility means more money.

The upshot of this is that the larger market volatility is, the more investors adopting a dumb strategy like that stand to make. More volatility means that time is being "compressed", with more swings per unit time and hence more money.

What I am saying is, this is a very basic trading strategy that means buy after a fall, sell after a rise. This is based on a random market with no inside information. In this, in so far as they can control it, there is no incentive for traders to keep volatility down. This would tangentially point to the conclusion that Finance, as an industry, has no incentive to do away with crises as those provide the biggest swings, and the way back up is paved in (bonus) gold.

Thursday, April 21, 2011

The Possibility and Case for a Public Insurance Scheme

It may be a massive cannibalization of industry, but there appears to be a case for public insurance. This became clear to me when I ran some simple numbers.

My insurance agent showed me a plan where a term policy with a $1M sum assured would cost over $200 per month. I thought it was outrageous (i.e.: working on an inflated probability of death), so I went home and did some estimates. (Point of separate note: NTUC income has a term policy with a $1M sum assured for 20yr for $60 per month given 30 was one's last birthday.)

SPF statistics reported about 285 fatal/serious accidents annually for 2006 and 2007 when Singapore's population was about 4.5M. This gives an estimate a lower bound for the rate of death and serious accidents for all causes each day (1.73 x 10-7). Let's suppose, to be conservative, that an upper bound for the true rate is 3 times that since reputedly road accidents are a "leading cause of death, et. al.".

Here is a simple model for insurance. Each day an insured party pays a day's premium, if he/she encounters an unfortunate event (with the above probability), a payout is made and premium payments stop. If nothing happens by a terminal age of the insurance policy (I used 40 years), no further premiums are paid and no payouts are made (dealings with the insurer end).

Going into some technicalities (skip the paragraph to get to the conclusion), the geometric distribution was (obviously) used as a basis for the computation of the expected income of the agency. The answer comes is three parts: (a) premiums paid given an incident, (b) incident payout given an incident, (c) premiums paid given no incident. An annoying thing happened to me when I was calculating this. I was evaluating an expression (I think it was (a)) which was the derivative of some expression that I knew to be increasing. It wasn't, at least until I realized that Excel's LOG function is base 10. (One performing this computation with some "numerical sense", will know why I bring in the LOG function.)

In any event, it turns out that my insurer uses at least 91% of the premiums to cover overheads such as commissions. NTUC uses at least 73%, noting the lower premium. This is before factoring in the higher premiums at later ages. This is high.

Given a buffer that can be tapped into and repaid slowly with low interest, an insurer can insure 4M citizens for 70 years from birth with a $1M sum assured, working with $352M annually for administration and accumulating $89M on average each year for a $25 monthly premium. The buffer is there to handle variance in payouts between years. One would expect about $761M in payouts each year, and assume that 99% of the time payouts are less than $1.52B, within 9 years one would accumulate enough to buffer against a $1.52B payout year without tapping the buffer. Within 20 years, the scheme would probably never have to touch the buffer. Once a sufficiently large reserve is accumulated, surpluses could then start going into a foundation to provide opportunities for the under privileged or some other worthy cause.

$352M administration for 2M citizens may be low. But this works if business expenses are kept low. Especially marketing and sales. Also, the buffer may be operationally realized, initially, by government debt via the bond market. Debt can be rolled over the same way most governments and companies do. The national reserves do not have to be touched at all.

As a start such a scheme might just handle only accidental death and Total Permanent Disability (TPD) up to age 70 (at which age it might be assumed that the formerly insured would have been able to provide for his/her dependents). This would ensure that the low administration cost would be easily achieved. Perhaps even more rainy day reserves could be accumulated if the costs turn out to be lower.

There are issues to be worked out. Such as costs for dealing with insurance fraud. Incentives for fraud can be handled by dynamically adjusting the sum assured to annual income based on the last tax assessment (and charging premiums accordingly), and perhaps with a tapering of payouts as age 70 approaches. Payouts could be made gradually to lessen the effects of variance. However, these are problems in incentives and operations that can be worked out.

This is heavy duty security for our citizens. Who cares if we drive the death and TPD wings of insurance agencies out of business. They weren't using much of our money to protect us anyway.

If this goes through, a national healthcare system might be possible. I apologize to the L-p norm loving people (for small p), but I take a "uniform convergence" (L-infinity) view of approach to a first world nation. Rather than just increasing the total welfare, we have to make sure that the least well off are lifted up.

This should be an element of our national strategy -- the elimination of accidental poverty. Notably, this does not entail transfer of wealth from the rich to poor. It is just a matter of good management and economies of scale.

Singapore has already disadvantaged singles as a side effect of part of its national strategy. My sense is the government has been too friendly with big business to the detriment of many of its citizens (i.e.: when interests are not weighted by income). Why should part of the insurance industry be sacrosanct?

Monday, April 18, 2011

On Housing: Asset Enhancement is an Illusion

The topic of property prices and their impact on both the young and society at large has been talked about recently ('Perils of asset enhancement for younger Singaporeans', 'Perils of asset enhancement for younger Singaporeans', ST Forum Apr 18). I'd like to weigh in on this matter.

A quick calculation in Excel based gives the estimate that a dual-income household with husband and wife beginning with somewhat above the median income ($3000, to simulate the time taken to build up 10% of the home value) would take about 10 years to fully own a $300,000 HDB flat. This is based on a 3% annual salary increment, their use of 30% of their take-home pay to service their loan, and an interest rate of 2.5% (based on POSB Home Ideal). Ten years is a long time. Furthermore, if the trend of property prices rising faster than incomes continues, repayment periods will get even longer. This points to a trend of increasing financial hardship, with families having less of their take-home pay available to them for longer periods.

Furthermore, assuming that a home is a necessity to a family, the family's assets are not really appreciating as other home prices are rising in tandem. Only when the family acquires a second home can it be truly said that their assets are appreciating in value. For those who do not yet own property, "asset enhancement" is a form of inflation -- the vaunted "taxation without legislation" which "greases the wheels of the labour market"; for those who own a single apartment/house, "asset enhancement" matters little; for those who own more than one apartment/house, "asset enhancement" is truly asset enhancement. With this laid out, "asset enhancement" is structurally similar to a regressive tax, which is contrary to the stated intention behind it.

Thursday, April 14, 2011

On National Strategy

Over the past years, issues of housing, inflation and real household income growth/decline have been increasingly central in the public discourse, pointing to the deeper issue of what national strategy is appropriate going forward.

Traditionally, we have pursued a high rate of growth which has, unfortunately, been accompanied with a rising cost of living relative to income, which in turn has led to many Singaporeans feeling left behind. While one must acknowledge that billions are spent on targeted aid to speed up the trickle-down process, which usually takes decades in other countries, targeted aid is a reactive solution to a side-effect of our pursuit of growth.

Our first generation leaders presciently pursued the strategy of growth while building an umbrella of security. The execution of this strategy raised the average standard of living and kept Singaporeans safe, and is now well studied by both governments and corporations as a textbook case study of well-crafted and implemented national strategy.

Empirically, economists have observed that high growth in a country is often accompanied by rising prices and increasing concentration of income with the richest, in turn leading to decreasing purchasing power of the poorer segments of that country. This leads to the question of whether high growth and increasing income inequality is, on balance, the best option for Singapore. In a related sense, is the growth chasing hot money a risk to our financial system? We have to make an informed decision on how to pursue growth and to what end. After all, the pursuit of growth should be a means to increasing the welfare of the citizenry.

Key to designing an appropriate national strategy is the articulation of our national objectives in terms meaningful to Singaporeans. Such objectives might include ensuring, by 2020, than no more than 1% of adults in a certain age bracket fall below a set income threshold (as pegged to a representative basket of goods and services), which, under the CPF scheme, would ensure sufficient retirement funds for most Singaporeans. National objectives might also include a list of privileges we would like Singapore citizens to enjoy, such as substantially longer paternity leave with job security or free tertiary education. Such examples represent a consolidation of the gains that generations of Singaporeans have labored for. With a clear sense of what we want to achieve, Singapore will be better able to negotiate the trade-offs necessary to realize her objectives.

As PM Lee as emphasized, this coming election is about the future. As such, and our nation’s long term objectives and strategy should be debated.