How to Save Social Security, take II!
I was reminded today of the importance of making policy change proposals simple and compelling. So I decided to rewrite my earlier idea. The basic idea would be to guarantee a 6% return for everyone’s savings held in the Social Security Trust Fund (SSTF), by having the US Government set up its very own mutual fund, US Mutual, using just a portion of the SSTF and a conservative nearly completely computerized investing strategy based on a simple algorithm using 27 weeks of NYSE stock-market data.
The rest of the idea is at my main blog here.
dlw
Advertisement
I may get more ambitious in my ideas for Stock Market reform. I think there's something to the notion that the log-return is a better way to judge stocks than the return, where log-return=ln(1+return). It might be a good idea to mandate papers or funds to share with people statistics like what US Mutual would be using to determine its stock-holdings, instead of the current statistics that they list.
I also think it matters that many current mutual funds perform not that well relative to the overall average. If the NYSE were less volatile due to the presence and the investment strategies of US Mutual, then I'm not sure I'd see any objective point for us allowing for private mutual funds to operate. The US Gov't could set up a public-run duopoly in the mutual fund business. The second fund would use the same basic approach as the one described above but with an index of value that would weight the predicted median log-return twice as much as its predicted median standard deviation. People would then be allowed to either pick their stock investments, individually or as part of a group, or buy into either fund or a combination of the two. We could even allow risk-loving investors to combine buying long in the less conservative US Mutual fund with selling short in the more conservative one or for more risk-averse investors to do the opposite.
Mathematically-speaking, you really only need two mutual funds with different investment strategies to accomodate a wide range of investment preferences. I think it would be to the advantage of the individual investor if there were only two mutual funds, and both used economies of scale and a simple algorithm to pick stocks relatively conservatively in a manner that helped to reduce the overall volatility of the NYSE.
This would also promote better stewardship of our economy by attracting capital away from hedges and helping the smaller and medium-sized firms to raise their needed capital. It's common knowledge that such firms are often better at job-production in the first place and would be worthy of our support.
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 8, 2006 12:56 AM | Reply | Permalink
I've thought about ideas similar to yours, but I am not persuaded SS needs saving--that is, I don't see the threat as either imminent or dire.
Also, all investment-type alternatives are not a replacement but a dramatic change away from an insurance policy. As an employee expecting a defined-benefit pension I would not accept an investment as equivalent.
If we can afford to go into debt by something approaching a trillion bucks for no return other than dead people (Iraq) I think we can afford to fund the insurance policy known as SS. Last I heard, the projection showing SS dropping below break-even in 2042 is the mid-level one, better than the worst-case and worse than the best-case. Previous projections, as I understand it, have tended to match best-case scenarios by SSA, in which case there is no problem, period.
July 8, 2006 9:23 AM | Reply | Permalink
Well, SS may not need "saving", but my approach would allow us to guarantee a better return for everyone's SS savings.
I don't think my idea means giving people the share in US Mutual. People could buy into US Mutual on their own. But everyone would receive a guaranteed 6% return on their savings, and those with less savings would do somewhat better in accordance with existing redistributional policies...
I'd say the idea is less about saving Social Security and more about stabilizing the NYSE and making it thereby more friendly for individual investors and small-to-medium sized companies for raising capital.
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 8, 2006 10:08 AM | Reply | Permalink
I am flexible in how the Social Security funds would get paid out, whether it is like a defined-benefit pension or a 6%( or more) guarantee on people's savings. My idea is more about both accelerating the growth of the SSTF and stabilizing the NYSE.
I think my approach would also not rely on illegal immigrants paying in with no hope of getting any return. I generally think that we should have them as guest workers and that, in return for registering and agreeing to pay higher income taxes and agreeing to improve their English skills as a condition for the right to work in the US, they would get legal protections(like the right to collect what they pay into Social Security), some free initial English training, transportation back to Mexico in case they fail to learn English adequately(at the expense of their own gov't) and, perhaps, some assistance with their future participation in Mexican elections and the long-term possibility of a joint-citizenship if they work hard and stay out of trouble with the law.
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 8, 2006 11:55 AM | Reply | Permalink
How can you say this? Have you never heard or read the ubiquitous disclaimer in the investment community?
Every investment firm has compliance officers who read all out-going correspondence, all prospectuses, all research reports, in short, everything to make sure no employee guarantees any ROI.
Why do you think they do that? Because they will have to pay off on the guarantee if the investment fails. Are you personally prepared, or even able, to guarantee a specific ROI to every retiree?
July 8, 2006 11:29 AM | Reply | Permalink
Firms need that saying to avoid getting sued in the event of a loss.
Have you ever heard of the notions of economies of scale? US Mutual would have leverage and add stability to the NYSE. It would only buy long on the more stable stocks and hedge with selling short on unstable underperforming stocks.
There may be few guarantees but it is an objective reality that the combination of both economies of scale and a wide diversification and low overhead will lead to a stable well-performing mutual fund that will allow us to increase significantly the guaranteed return to people's Social Security savings.
If we can stabilize the Stock Market, we needn't preoccupy ourselves with worse-case scenarios that stem from past irresponsible speculations.
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 8, 2006 11:41 AM | Reply | Permalink
Since you were nice enough to explain a bt of the theory in email, I should explain my disclaimer -- I do understand the statistical methods you describe, but I don't begin to pretend I have a real understanding of market econometrics. Network protocol theory and medicine, yes, and even a bit of rocket science.
From my experience working with economists, you can't really be one until you get proficient in sacrificing goats and chickens.
--
Howard
*equal opportunity offense to both extremes*
July 8, 2006 12:26 PM | Reply | Permalink
Well, if you are familiar with the concept of median, that shd suffice to explain why the algorithm would be stable over time.
The biggest problem with having a large public mutual fund is its potential for corruption. If the holdings are based on relatively simple statistics using publicly available data then that potential for corruption would be circumvented.
Once I get some real stock market data, I'll be able to demonstrate further with relatively simple statistics why my idea would make for an ideal way to conservatively invest the SSTF so as to provide for baby-boomers like yourself when you retire...
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 8, 2006 2:25 PM | Reply | Permalink
I am, indeed, familiar with medians, just as I am familiar with other things you have mentioned, such as kurtosis.
On a more or less weekly basis, I work with telephone and Internet traffic engineering. To get a good handle on demand, one certainly must deal with things such as time series variation, trend smoothing, etc. In general, I know these well enough that I can use them as inputs to decent statistical packages, or, when necessary, call in a statistical consultant.
CIA analysis of the Soviet economy was not at all ideal, but was certainly better than nothing. A large part of their work was econometric modeling, and a good deal of this was declassified. It wasn't trivial, although there was -- and this was years back -- a pretty nice, general purpose time series analyzer that I used for computer capacity planning. I just am not convinced simple statistics will be adequate here.
--
Howard
*equal opportunity offense to both extremes*
July 8, 2006 3:19 PM | Reply | Permalink
If the median log-return outperforms other measures on historical data in prediction then it and it's median standard deviation shd suffice for ranking and picking stocks.
When I was an economics professor in Mexico I taught econometrics. We had a large number of data sets and I always emphasized my students getting their hands dirty on real data over the mathematics. I also emphasized that creativity is also important. We regressed weekly returns of the NYSE index on one its lag values. If you ran a mean regression then the correlation on the lagged value was statistically insignificant and it was pronounced that the market was efficient and that it was impossible to time it. But when you ran the same regression using a median regression, the correlation was statistically significant and I think also economically significant. What was mucking up the other regression was the inclusion of the 87 stock market crash outlier.
I also helped to supervise a thesis on the Mexican Bolsa Stock Market and found that the median autoregressive regression both had a significant statistical correlation and predicted better than the mean version over the course of a three or four week horizon.
It's not an exact science, but it doesn't need to be. It just needs to be transparent, stable and somewhat flexible, so as to change the underlying behavior of the NYSE and eventually that of other stock markets...
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 8, 2006 4:30 PM | Reply | Permalink
High performance is what we seek by risking investment funds. Reliablity is another animal altogether.
For me the value of the SS income I expect is not its size but its certainty. Why does anyone buy insurance? Sometimes certainty outweighs all other considerations. If my car is hit by another driver's, I care not at all whether his fund has deep pockets, only that it has enough.
I haven't seen much discussion, (except maybe by Krugman), of the effect of a large segment of the population liquidating or restructuring assets simultaneously, as would be the case when a demographic bubble hits retirement. First approximation suggests values dropping across the board. Then there would be a run on the funds to save value for the early sellers.
July 8, 2006 5:55 PM | Reply | Permalink
The two are not completely incompatible.
We wouldn't invest the entirety of the SSTF in US Mutual, we'd only target a 6% control of the total predicted stock market valuations. The more we grew the SSTF in the near future, the easier it would be during the transition and it would also help if we made the rules change so that immigrants or guest-workers would get to contribute and benefit from US Mutual.
Trust me, if we offered something closer to a 6% guaranteed return on people's SS savings, we'd probably get a lot of takers that would keep the system afloat.
dlw
A blog-activist dedicated to the reduction of the faith-based political acrimony in the United States of America so as to make our political system more democratic and just and to improve our witness to the rest of the world.
July 9, 2006 12:57 AM | Reply | Permalink