Bogleheads.org

There was a time due to the mechanics of the stock market when dividends made sense for retail investors. It was only 25 or so years ago when discount brokers started becoming popular. Before the 90s, a full service broker charged over $100 just to make a trade and many even based the fee on the number of shares transacted. Dividends and dividend reinvestment plans (DRIPS) provided a convenient means for people to access the revenue of a company and grow their stock position while bypassing the “Wall Street tax”. Also, shares weren’t traded in pennies, but other fractions of a dollar and you could only trade in full shares with a bias towards round lots of 100 shares. That could be a hurdle for retail investors. Even when discount brokers c...

I forget where I read it. It was years ago, but it was to the effect that people that claim at 62 receive on average the same benefit in total dollars received as those that delay until 70. They receive a smaller amount over a longer time frame. Those that delay have higher monthly payments but for a shorter time frame.

Social security is more than just a math optimization problem. Personal and family health history play a big part of the decision.
Though the population is getting older on average, plenty of people still die before reaching full retirement age; let alone 70.

But I do think they are rules based and have no individual security selection or market timing. Help me understand something please. Yes, quant funds have algorithms and rules-based processes that allow them to trade to a consistent strategy, but those human-built algorithms still produce buys and sells of individual securities, don’t they? I don’t see how the “human or algorithm” question relates to the “passive or active” question. It seems like the quant guys still are picking individual stocks as opposed to buying the whole basket. Not true? What is an index but a human designed algorithm for selection? They may be very simple algorithms (top 500 largest US companies weighted by market cap), but they are still algorithms. Portfolio con...

Gaston wrote: ↑Thu Mar 02, 2023 4:05 pm

Random Walker wrote: ↑Wed Mar 01, 2023 9:37 pm But I do think they are rules based and have no individual security selection or market timing.

Help me understand something please. Yes, quant funds have algorithms and rules-based processes that allow them to trade to a consistent strategy, but those human-built algorithms still produce buys and sells of individual securities, don’t they? I don’t see how the “human or algorithm” question relates to the “passive or active” question.

It seems like the quant guys still are picking individual stocks as opposed to buying the whole basket. Not true?

What is an index but a human designed algorithm for selection? They may be very simple algorithms (top 500 largest US companies weighted by market cap), but they are still algorithms.

Here are some interesting comparisons.

Portfolio Visualizer: PSLDX.

An 80/20 psldx/cash mix out performs 100% spy with similar risk to an 80/20 stock/bond portfolio when rebalanced monthly. It does so with less volatility and draw down than 100% psldx (not surprisingly).

The benefits of diversification are undeniable. The holy grail is to have two asset classes with high returns and zero correlation. But look at the history: - International stocks used to be a good diversifier. It became widely available, a lot of institutions started investing into it. Whoops, it's now strongly correlated. - Real Estate used to be a good diversifier. Then REITs came along, a lot of investors started buying them. Whoops, it's now strongly correlated. - Commodities used to be a good diversifier. Too many people realized that and rushed into buying them. Whoops, it's now correlated and returns are no longer great. It seems that whatever high return (or even medium return) asset class becomes widely available, it very quickly...

Don't base investing decisions on current borrow rates. Theoretically the return of stocks is higher than whatever the risk free rate is (the borrowing rate). Which is called the equity risk premium. You also have to consider the sequence of return risk by playing market timing games The market could go up 25% next year. I guess this also means we can consider using margin to acquire the equity portion in my taxable account if needed. I assume it's never a good idea to buy bonds with margin (over LETFs)? Yeah we don't buy bonds with margin/box spread because the borrow rate is a hair higher than what you get implicitly in the treasury future. Also while IB would let you lever actual bonds heavily, the margin requirement of bond etfs is hig...

I do returns also and I agree with the prior comments. Additional thoughts. MFS can put over 65 taxpayers in higher Medicare IRRMA rates. In CT where I practice, many times MFS saves CT tax while federal stays the same. Taxpayers with student loans sometimes do better on their payment requirements when filing separate. Couples that are both working should leave their withholding using single/MFS rates on their W-4s so they don't underpay even if filing jointly. When filing separate need to be careful about matching tax payments and withholding going forward. If a balance is due make sure the spouse who doesn't owe if filing separate knows they do not need to file jointly. Many believe they have to file joint. Filing separate is the default...

401k plans have a lot of regulations that must be adhered to, particularly with highly compensated employees. I've not heard of any 401k plan that allows transfers out while employed because it would interfere with the metrics that plans have to track in order for the 401k plan to be compliant.

I don't know if HSAs have similar requirements.

Follow up ...

I found a python package that is a very thin wrapper around the treasury direct api.
https://github.com/npezolano/PyTreasuryDirect

It does not provide price data, but it's a start and shows some usage of the api.

Ah. Didn’t see the takedown notice. Thanks

Got an error page when accessing via Safari on iPad.

Dry-Drink wrote: ↑Sun Jan 08, 2023 1:37 pm

ToBeOrNot wrote: ↑Sun Jan 08, 2023 12:29 am And while there isn’t a circuit beaker supporting TLT, the underlying bonds aren’t that volatile to move 33% in a single day (I may be misremembering some details on TLT/TMF).

NYSE does have a circuit breaker for TLT as part of its Limit Up Limit Down program.

Thanks for the correction. I wasn’t quite sure.

So then TMF won’t blow up either in a single day.

Because of the structure of UPRO, talking about anything except daily returns is unreasonable. There is not a linear relationship between UPRO and SPX. Even in the chart posted of SPY and UPRO for 2022, notice they are not linearly correlated. For a 25% draw down in SPY, UPRO did not fall 75% but rather a bit over 60%. Volatility decay provided a small cushion. Also, talking about individual components isn’t reasonable either. It’s the combination of uncorrelated volatile components in a frequently rebalanced portfolio that’s key to the strategy. But none of this is new. It is all covered extensively in the various HFEA related threads. Many people spent a lot of time not only looking at the starry eyed upside returns, but all of the pitfal...

As other posters responded to your original post, MPT does not rely on negative correlations. It relies on correlations less than 1. It does not rely on a 5 or 10 year backtest, but is supported by 60+ years of backtesting. A couple bad years is noise just like it is for the market, but amplified by leverage. An investors leverage should be based on risk tolerance with consideration of lifecycle investing principles. Personally, I own considerably more shares of sp500 than I did a year ago. A couple bad years leading to -33% stock market, -33% LT bonds means your portfolio is almost 0, only saved by leverage resets. Even worse with the gross expense ratios being paid. Have you read all of the related HFEA threads? Whether one agrees with t...

That sounds like a made up hand-waving argument to me. I'm a fundamentalist, so explain it to me about stagflation as if I were a fundamentalist. I don’t know what you mean by fundamentalist (could mean multiple things) so I don’t know how to define stagflation for you. Are you unfamiliar with the economic conditions of the 1970s? Among many characteristics, ( Great Britton had similar conditions as well) the economy overall stagnated (stag-). There was rampant unemployment, markets were falling. The dollar was unpegged from gold. And yet despite poor economic conditions, instead of deflationary pressures, there was run away inflation (-flation). It wasn’t until 1981 that the FED with Volker at the helm started to aggressively fight inflat...

Raising interest rates isn’t the problem, even in a falling market. It is a 70s style stagflation that kills this strategy. The belief was that since Volker showed how to tackle that, we wouldn’t face it again in the future. The strategy backtested ok during the Greenspan era of rising interest rates because inflation was kept relatively in check and bonds still served as the flight to safety after the dot-com crash. Unfortunately, the FED and the current administration kept saying (wishing) last year that inflation was only mild and temporary and took too long to start raising rates. The strategy seems to be performing exactly as expected relative to the 70s in a stagflation like environment. We don’t know how long inflation will persist. ...

Thanks for the info.

I was also able to find https://www.nyse.com/products/bonds. You can get a csv of some listed bonds, mostly corporates.

In addition I found www.cbonds.com. They provide api access but charge for data. Which is understandable because curating data can be expensive. Not sure how much they charge though. They only provide a “request quote” link.

For equities, there are plenty of places to get both current and historical price data that are easy to access via api calls from your favorite programming language. Many of them are even freely available if you are ok with end-of-day quotes or delayed quotes. Do similar sources exist for bonds? Treasuries, Agency bonds, munis, etc?

I ask because I’ve been looking for bond ladder tools. Fidelity has a particular nice tool, but it is limited to 50 rungs and has other limitations. Others has similar limitations. So I was looking for data sources that I could access via python. Are there any readily available?

We have to remember that the 20% loss is baked into the average market return. . That means that some future years will have above average returns. Just because there is a $15k loss today, you are not doomed to missing out of $58k of future returns at 7%. Some years will have 25% (which makes up for the 20% shortfall) or greater return. As others have pointed out, the SP500 has had a 50% drawdown in the past. That loss is also baked into the average return of 7-8%. This argument appeals to the "reversion to mean" model as opposed to "random walk" model - discussed only recently here: https://www.bogleheads.org/forum/viewtopic.php?t=392760 Some future year isn't going to have higher return just because this year the mark...

We have to remember that the 20% loss is baked into the average market return.. That means that some future years will have above average returns. Just because there is a $15k loss today, you are not doomed to missing out of $58k of future returns at 7%. Some years will have 25% (which makes up for the 20% shortfall) or greater return. As others have pointed out, the SP500 has had a 50% drawdown in the past. That loss is also baked into the average return of 7-8%.

Definitely need to make sure to use European style options. Using American options is how that poor soul bragging about box spreads blew up their account. Market Watch: 1RONYMAN.

So in the backtests, when did the strategy zoom into the millions of dollars? Was it gradual or a specific time of falling rates and rising market like 80s-90s? If I remember, one backtested period resulted in high volatility and similar returns to no leverage. I do not know which backtest you were looking at, but I suspect it didn't have the data from 1965~1980. This was when this strategy absolutely would've gotten rekt had you been pouring money into it. I recall simulated tests using data going back to the 50s in the original thread. There was a whole other thread where data generation was discussed. And no, HFEA did not perform well in the 70s. That’s where the 60-70% draw downs occurred. That time period was also used as justificatio...

So in the backtests, when did the strategy zoom into the millions of dollars? Was it gradual or a specific time of falling rates and rising market like 80s-90s? If I remember, one backtested period resulted in high volatility and similar returns to no leverage. I do not know which backtest you were looking at, but I suspect it didn't have the data from 1965~1980. This was when this strategy absolutely would've gotten rekt had you been pouring money into it. I recall simulated tests using data going back to the 50s in the original thread. There was a whole other thread where data generation was discussed. And no, HFEA did not perform well in the 70s. That’s where the 60-70% draw downs occurred. That time period was also used as justificatio...

To my knowledge the account has to be open for 5 years. There is not a limitation on the funds in the account once the account meets the 5 year limit and the owner is at least 59 1/2 years old. Principle can be touched without limitations because taxes have already been paid.

Settlement times are a legacy from when it took days to complete and settle stock trades.

One way to get around that is with a margin account. Many brokers today even offer cash settled margin account in retirement accounts. You won’t be able to use margin loans to leverage up in retirement accounts, but you can trade immediately before settlement.

When playing around with “Profit Farmer” variations of the strategy on portfolio visualizer, I found that draw downs and portfolio volatility were very reasonably contained without sacrificing too much CAGR when applying a 5% portfolio stop loss. This year has still been a series of losses but not quite as bloody. How is that stop loss implemented, seems like many days are over 5% +/- with a strategy like this, is it stopped on on daily volatility or monthly, and what is the mechanism for getting back in once stopped out? It looks like “sell at the close” for a trading day that has total portfolio value lower than 5%. Or you could sell at the next open. Then restart the strategy on the next rebalance day next month. Yes that means you coul...

When playing around with “Profit Farmer” variations of the strategy on portfolio visualizer, I found that draw downs and portfolio volatility were very reasonably contained without sacrificing too much CAGR when applying a 5% portfolio stop loss. This year has still been a series of losses but not quite as bloody.

I’m in the LITE commissions pricing structure and haven’t made a trade in over 18 months and haven’t had a threat of shutdown. I do however have periodic deposit activity.

Check which price structure your account is configured for. The LITE model doesn’t have inactivity fees.

Interactive Brokers allows you to borrow at just above LIBOR against your invested balance. You could put your savings into treasuries there for yield and borrow against yourself if you need it. You wouldn’t have to ladder for availability, but you could for yields since interest rates are rising.

Interactive Brokers Cash Management

Just for kicks, I analyzed the performance of 55% UPRO and 45% TMF since February of 2019 , the month when HFEA was first posted. Through the end of April of this year, the S&P 500 averaged 1.27% higher annualized returns and had FAR lower volatility and maximum drawdown. You forgot to rebalance quarterly. I didn't recall that being part of the strategy. Rebalancing quarterly resulted in much higher returns, but rebalancing monthly resulted in returns between annual and quarterly rebalancing. I smell a fluke, and I don't mean fish. Go back and reread the first HFEA page. Quarterly rebalancing is crucial to the strategy. Some found that monthly rebalancing with inverse volatility weighting had a slight CAGR advantage. That was based on ...

The fundamental broad assumptions of the strategy are 1) bonds and stocks have returns greater than the cost of borrowing 2) bonds and stocks have a correlation less than 1.0000000000. 3) the combined portfolio has low enough volatility to lever to 3x and/or you make contributions from income that reduce the risk of volatility decay I don't see any evidence presented in this thread that those assumptions have changed. In the mHFEA thread we have questioned #1 somewhat. Specifically whether bonds have returns greater than the cost of borrowing or a correlation with equities negative enough to justify returns near the cost of borrowing. If you get into the specifics, it assumes 55/45 is the best ratio, the LTT are preferable to ITT or anothe...

In mathematics, even if the probability of a particular event is zero, it's still possible for the event to occur! It is by definition not possible for an event to happen if the probability of the event is 0% with no margin of error. Perhaps you are trying to say that we can look at something and say it has happened 0 times out of N samples, therefore we calculate the probability of this happening to be 0%, but in reality the real probability isn't 0%--its higher, we just don't have a large enough sample size yet? Sorry but this is not correct. An event of 0 probability can still happen. The event is just highly improbable, but not impossible. Likewise an event with a probability of 1 may never occur. That event would be highly probable, b...

This thread and both of the original HFEA threads have been quite interesting. Oh, and Hello Boggleheads! Using the portfoliovisualizer links at the start of this thread, I noticed some interesting behavior in various back tests. 1. Returns seem to be driven primarily when equity allocation is between 125% and 165%. 2. ITT provides very interesting results when ITT is between 2x - 6x equity allocation. 3. With ITT at higher multiples, while overall portfolio stdev was higher, both maximum draw down and worst performing year were smaller than 100% SP500. Returns were also orders of magnitude larger than any other similar portfolio, including original HFEA. I've been running mHFEA in a paper trading account at IBKR. I've got some questions f...

This thread and both of the original HFEA threads have been quite interesting. Oh, and Hello Boggleheads! Using the portfoliovisualizer links at the start of this thread, I noticed some interesting behavior in various back tests. 1. Returns seem to be driven primarily when equity allocation is between 125% and 165%. 2. ITT provides very interesting results when ITT is between 2x - 6x equity allocation. 3. With ITT at higher multiples, while overall portfolio stdev was higher, both maximum draw down and worst performing year were smaller than 100% SP500. Returns were also orders of magnitude larger than any other similar portfolio, including original HFEA. I've been running mHFEA in a paper trading account at IBKR. I've got some questions fo...