
Discover more from Known Unknowns
Allison's ode to the second moment
Hello,
Welcome to the 31st issue of Allison’s Ode to the Second Moment, a newsletter where friends don’t let friends invest their retirement savings in crypto-currencies.
I try not to judge, but I think this is nuts
Did you know you can invest your IRA in Bitcoin or Ethereum? What’s even crazier is that almost everyone I speak to thinks this a great idea. After all, Bitcoin prices are up more than 300% in the last five years. There is a finite supply of them. Prices can only go higher, right?
Uh sure, or they might crash if people decide one day that a made-up currency that's not backed by any government is worthless. Such a high return comes with considerable risk. Monthly volatility is more than 90%!
The Bitcoin-IRA people explained to me that investing your IRA in Bitcoin is a great idea, in part, because it provides a good hedge against the dollar. Sure. But our retirement consumption is spent in dollars. I am not sure why (unless you seriously fear hyperinflation) future retirees need to hedge against movements in the dollar.
This is what happens when people think about investing in wealth rather than income terms.
On the other hand, how else will we pay for long-term care?
Based on the new budget, it seems the GOP plans to make some cuts to Medicaid. Some argue that cutting Medicaid turns America into Gilead. Arguably, the elderly would be harmed the most. Americans over age 65 constitute just 9% of Medicaid recipients, but account for 21% of Medicaid spending. Medicaid pays for many people’s nursing home care. Medicare does not cover long-term care, but if you qualify, Medicaid will pay for it.
This creates all sorts of undesirable incentives. Most elderly people need the state’s help. Others benefit by spending or gifting their assets, then claiming Medicaid to pay for their long-term care. This will only get worse. About one third of people over age 85 have some form of dementia. As people live longer, many will require long-term care: a large share of that burden will fall on Medicaid.
This is not a problem unique to America. Theresa May made a rather bold proposal that long-term care would be paid out of people’s estates after they die. If you don’t end up needing long-term care, you don’t pay the tax. This is referred to affectionately as the “dementia tax.” The proposal had some problems. However, the general idea was not terrible.
As you can imagine, the whole idea was wildly unpopular. The proposal was abandoned. The problem is not going away. A greater proportion of the population will soon need long-term care, someone has to pay for it. We need to have an honest conversation about who that will be.
Speaking of being in the dark about who pays for what, Buttonwood is especially great this week. He explores the erroneous belief that we've paid for our entitlements.
Robots are coming for pension geeks too
Risk assessment has always been a tricky business. Regardless of how much data we seem to have, it always comes down to human judgment—or error. Determining what data to use, never mind the techniques we use to estimate risk, takes knowhow.
This situation may be changing. Big data/machine learning are transforming the risk models insurance companies use. They not only provide better data on risk factors—insurance companies can monitor customers to police moral hazard! It’s like 1984 for risk nerds.
AIG pioneered some exciting new techniques that have the potential to price risk with greater accuracy. They brought more science to risk assessment. Thus far, the market has not appreciated the effort—good risk management is rarely recognized when times are good. Some analysts are not fans either; they are reluctant to cede their skills to smart machines. They insist that policies are too quirky to be measured with data alone. Looks like their jobs are safe for now.
The same cannot be said for former Ivy League jocks.
More evidence that technology benefits the geeks
Uber has been practicing price discrimination. They charge prices based on how much they think we can or will pay. For example, someone traveling between two expensive places might pay $10, while one traveling between two low-cost places (at the same distance) will pay only $3. Uber also is not cutting the drivers in on the higher prices. Drivers receive the same pay for both customers. This all sounds terrible to some. To be fair, Uber does not have a great track record when it comes to public relations.
I am not sure how prices are set exactly. In principle, price discrimination isn’t necessarily horrible. It may result in lower Uber prices for people in low-income neighborhoods. Remember, Uber filled a large void for under-served neighborhoods. If drivers get different fares depending on the client’s ability to pay, they might refuse lower income customers. For price discrimination to work, Uber might need to pay drivers a standard rate. If that rate is high enough—drivers might be better off because they have more customers with less risk.
What everyone saw coming: Municipalities in trouble
More sad news from Puerto Rico: its largest public pension fund ran out of money. So begins the battle between pensioners versus bondholders, the very same people who are funding the pensions.
Things are only a bit better on the mainland. The Wall Street Journal declared Connecticut has run out of rich people to tax to pay its obligations.
That’s a good reason to account for risk in your discount rates—am I right?
Pension Geeks 5000
Actuaries -1,000,000
Speaking of pension shenanigans, you know how one of the advantages of DB plans is controlling workers' tenure? It turns out that teacher pensions operate under the assumption that the DB plan vesting rules have no influence on tenure.
I’d also like to point out this newsletter did not use the word “and.”
Until next time, Pension Geeks!
Allison