Valid, Not Optimal

A couple of things from Barron's. First an article about buying low and selling high not "working" during this event. There is a focus in the article about whether or not to reduce exposure to stocks doing well to rebalance into stocks that are struggling.

Included in Adam Parker's comments was to "only buy losers when you are confident that a market bottom is close at hand and a big recovery will follow." When you're confident the market has bottomed? That sounds pretty easy, amirite?

You'll find plenty of differing opinions about trimming winners to buy laggards but replace "confident" that the market has bottomed with add more net long exposure after a large decline fulling realizing you might be wrong for a while. Buying after a 20 or 30% decline won't be emotionally easy and yes you could absolutely be early on the way to a 40% decline but buying after large declines will work out far more often than not. 

Another article sought input from advisors about what they do to help clients avoid running out of money. One advisor laid out a strategy that "typically divides clients’ assets into five to six buckets based on time horizon and risk." 

Here are the buckets she uses and how she labels them;

• Cash for 1-2 year liquidity
• Short term stability for 3-5 years
• 60/40 for six-15 years (70/30 works too she said)
• Buffer fund for 15-20 years out
• Dividend stocks for 20-30 years out

What's your first reaction to that? Me too but backtesting her idea has a pretty good result.

Before any critique, here's what I used to try to replicate the advisor's strategy;

We'll get into some detail in a moment but her idea is clearly valid.

The 5 Bucket is not that far behind VBAIX in nominal terms and it has a better risk adjusted return as measured by the Sharpe Ratio. Several of the other portfolio stats are also superior. There was no info about what specifically she uses to build these buckets but there seems to be some overlap between the first two. Maybe she is using individual issues and really is differing the maturities between each of those two buckets. 

I would not expect buffer funds to capture all of the upside. That's not a problem going in as long as you realize that. Copilot says BJAN is the oldest buffer fund so I chose just to get a longer back test. Since it came out it has compounded at 11.56% versus 15.74% for the S&P 500 with about 3/4 of the volatility. Adjusting for growth rate versus volatility BJAN's return has been about the same as the S&P 500. BJAN has compounded 114 basis points better than VBAIX which is interesting but with a more volatility.

The suggestion of using a buffer fund for a long term bucket might be puzzling. I've seen one other advisor suggest this and being blunt, I don't know why this would make sense. Copilot said it really doesn't make sense other than for behavioral reasons like a lower tolerance for equity market volatility. Given how close BJAN is to VBAIX, the entire 60% (40% plus 20%) could be put into one or the other, BJAN or VBAIX, assuming confidence that BJAN could continue to be a close proxy for 60/40 but without interest rate risk. FWIW, the two have a 0.96 correlation to each other. 

The article had no info on how the buckets are weighted so I put 4% in bucket one and 8% in bucket 2 to both correspond to a 4% withdrawal rate. The other weightings are pretty much made up, thinking 40% into the 60/40 bucket would do most of the heavy lifting in terms of growth and sustainability. 

So now, let's see if we can improve the Original 5 Bucket Portfolio with some of the strategies we talk about regularly. 

Everything but SHRIX is in my ownership universe. We're replacing VBAIX with SPMO/SPHQ, ACWX and SHRIX. FLOT instead of SHY. And all-weather replaces the buffer. SCHD plus SPMO/SPHQ sort of brings in the quality/value/momentum idea we looked at the other day. 

Again, I will say it is valid. The slightly better CAGR from our version is nice but I would focus more on the volatility, drawdown numbers and other portfolio stats. Those are probably more sustainable than knowing whether it can continue to outperform. I think avoiding interest rate risk will help it outperform but that could easily turn out to be incorrect. 

I asked Copilot to critique both of them. Its initial reaction was;

I explained the context. It isolated the issues with buckets one and two in the original version, it liked FLOT a little more than SHY for example. BLNDX is way better than BJAN, it said BJAN is not "compounding machine." I would argue it is, just not like equities. My weighting to ACWX is too small to matter but I would push back that as part of a 60/40 bucket it is sized about right. It thought both versions had too much in SCHD but the attempt there was to just be true to the original suggestion about dividend stocks. 

Based on previous interactions it said that my version was "convexity-aware" but that it lacked any "fast convexity" like BTAL. My final query was to note that as I said above, I think they're both valid but not optimal and Copilot agreed with that assessment. 

One final hit. We've looked at a portfolio that was heavy on cash or cash proxies and light on TECL which is a Direxion 3x bullish fund that references to XLK. XLK is down 14% from its high while TECL is down 47%. If TECL tracked exactly for a longer period (that is not what it should be counted on to do) then you might expect it to be down 42%.

This backtest assumes buying TECL at its high.

Until it's rebalanced, if the broad market continues lower then this blend will probably be more defensive for now only having 15% notional long exposure not the 30% that it started with. Knowing when to rebalance would be pretty tough though. 

The information, analysis and opinions expressed herein reflect our judgment and opinions as of the date of writing and are subject to change at any time without notice. They are not intended to constitute legal, tax, securities or investment advice or a recommended course of action in any given situation.