Vanguard hedged global bond index vs. VLS as the bond element decreases

coyrls

aroominyork said:

So you are essentially taking VLS out of the equation and are pointing out that the bonds in VLS perform similarly to the IA Sterling Strategic Bond (SSB) sector, and the SSB sector performs differently to a hedged global bond index. Is that right?

No, I don't think that's what Bowlhead said.  He is saying you shouldn't use either index as equivalent to a 0% VLS, as the VLS bond allocations are from multiple bond indexes and VLS does not seek to track either index with their bond allocation.

aroominyork

OK, I'll re-word it. "... the bonds in VLS have performed similarly to IA Sterling...".

bowlhead99

aroominyork said:

So you are essentially taking VLS out of the equation ?

I'm not taking VLS out of the equation at all.

You couldn't understand the difference in 'steps' between the different levels of VLS from  100% equity down to 0% equity.

By simply doing the maths of multiplying out the returns, after seeing that  VLS 100% equity gave 53.8% over five years, you can see that every level of VLS was in line with expectation, if you use an assumption that their bond mix gave between 23.5 and 23.8% over the period.

Is it unreasonable to assume that their particular mix of bonds happened to deliver around 23.5 or 23.8% return? No it isn't, because the average for the IA sterling strategic bond sector (a sector which was invested in a mix of bond types either in sterling or hedged to sterling) gave about 23% and Jupiter was about 25%. So the idea that the Vanguard VLS bond mix of bond types either in sterling or hedged to sterling, gave a return in that sort of ball park, isn't implausible.

What is unreasonable is for you to plug the performance of a specific bond index (which doesn't represent the actual mix of bonds that Vanguard use in the VLS product) as your proxy for the mix of bonds that Vanguard use in the VLS product, and then scratch your head and say, "huh? why doesn't it work?". Simply, you haven't used the mix of bonds they used.

There is no telling whether sterling strategic bond sector will on average perform the same as VLS's particular mix of bonds over the next five or ten years or more. My comments were simply to note that your use of either 'Global bond index' or 'UK investment-grade corporate bond index' as a proxy for the particular mix of bond funds held by VLS would not give you the result they got, because they didn't use either of those funds as the entire component of their holdings.

aroominyork

It was poorly phrased about taking VLS out of the equation. I understood your 1231 post so apologies for making you repeat the point a few hours later. It was more that I was taking VLS out of the equation/question after your explanation, and moved on to wondering whether a global bond index is generally less volatile than the IA Strategic Bond Sector.

The reason for (not for the first time) trying to get under the skin of bonds is that I recently increased my bond allocation as I wanted to de-risk my portfolio and I am an equities/bonds/cash person (no gold or alternatives). I split my bonds allocation 35% to Jupiter Strategic Bond as a well performing and well managed mid-risk strategic bond fund; 35% to Royal London Short Duration Credit, which has given steady returns with low volatility; and 30% to Vanguard Global Bond Index (hedged), which gives some low cost broad diversification. Any views on that would be welcome. I am very unlikely to make calls on specific sectors so I am not really interested in whether some people think it’s a time to overweight gilts, underweight linkers, go for high yield etc. I’ve read some of dunston’s recent posts and realise he would not cut the pie the way I have, but I want to leave the decision making to fund managers and hold the index fund for the reason I described.

aroominyork

Remember simultaneous equations from school? By using two versions of VLS, say VLS40 and VLS80, you can isolate the bond element in VLS to compare with Vanguard's hedged global bond index. The calculations below look at the last five years when VLS40 returned 41.33%, VLS80 returned 61.73% and the bond index returned 17.80%. 

So the two VLS funds can be expressed (e= equities and b=bonds) as:

40e + 60b = 41.33

80e + 20b = 61.73 

Which can be rewritten as:

80e + 120b = 82.66

80e + 20b = 61.73

 

Subtract one from the other:

100b = 20.93

 

So the bond element of VLS returned 20.93% compared to the bond index returning 17.80%. 

Curiously when you calculate year by year and then compound it, VLS's bonds return 22.70%. I don't understand that difference - the mathematicians among you might. See this link - data drawn from HL yesterday.

Alexland

aroominyork said:

Remember simultaneous equations from school? By using two versions of VLS, say VLS40 and VLS80, you can isolate the bond element in VLS to compare with Vanguard's hedged global bond index.

I don't think that's valid as there is some automatic rebalancing going on within the fund and that be generating it's own performance depending on the asset allocation ratios.

Another_Saver

1. Upweighting the UK

2. Rebalancing - if you bought 80% of VLS 100 and 20% of a hypothetical VLS 0% equity, the bond weight will fall over the longer term, rebalancing adjusts for this so using simultaneous equations won't work.
3. UK investment grade bonds cannot be compared with the global bond index - the latter is mostly gov bonds, the former is mostly corporate bonds. They are different asset classes.

aroominyork

Another_Saver said:

1. Upweighting the UK

2. Rebalancing - if you bought 80% of VLS 100 and 20% of a hypothetical VLS 0% equity, the bond weight will fall over the longer term, rebalancing adjusts for this so using simultaneous equations won't work.
3. UK investment grade bonds cannot be compared with the global bond index - the latter is mostly gov bonds, the former is mostly corporate bonds. They are different asset classes.

I started this thread by asking why the bonds in VLS have performed better than a global bond index. Your first and third may contain part of the answer: have UK corporate bonds outperformed global govt bonds during recent years? 
I am not sure your rebalancing argument works: you can either auto-rebalance within VLS or manually rebalance between an equity and a bond fund. The outcome would be much the same.

Another_Saver

aroominyork said:

Another_Saver said:

1. Upweighting the UK

2. Rebalancing - if you bought 80% of VLS 100 and 20% of a hypothetical VLS 0% equity, the bond weight will fall over the longer term, rebalancing adjusts for this so using simultaneous equations won't work.
3. UK investment grade bonds cannot be compared with the global bond index - the latter is mostly gov bonds, the former is mostly corporate bonds. They are different asset classes.

I started this thread by asking why the bonds in VLS have performed better than a global bond index. Your first and third may contain part of the answer: have UK corporate bonds outperformed global govt bonds during recent years? 
I am not sure your rebalancing argument works: you can either auto-rebalance within VLS or manually rebalance between an equity and a bond fund. The outcome would be much the same.

/
If you picked a 5 year period and compared a manual 80/20 portfolio with no rebalancing and VLS80, vls has rebalanced quarterly 20x in the period, the 80/20 fund won't have and the equity allocation will normally have grown, maybe to 85%. That is why you cannot compare them lineally.
You have sought to isolate the return of a hypothetical VLS 0 by ignoring the effects of rebalancing, you would have to go through the historical precise allocations at the start and end of each rebalancing period and do your equation.
Also, if you go on trustnet and compare vanguard global bond with uk gov and UK investment grade you can see the UK has outperformed. Among other causes the maturity on UK debt tends to be higher than the global norm to reduce interest rate risk to the gov, hence the fall in interest rates over the period you're talking about has caused greater returns in the UK.
Also you need to use geometric equations not arithmetic so I don't even know if that can be worked out. I might try it out on paper later but on phone atm.

Thrugelmir

aroominyork said:

Another_Saver said:

1. Upweighting the UK

2. Rebalancing - if you bought 80% of VLS 100 and 20% of a hypothetical VLS 0% equity, the bond weight will fall over the longer term, rebalancing adjusts for this so using simultaneous equations won't work.
3. UK investment grade bonds cannot be compared with the global bond index - the latter is mostly gov bonds, the former is mostly corporate bonds. They are different asset classes.

have UK corporate bonds outperformed global govt bonds during recent years? 

BoE has been a net buyer of Corporate Bonds under the various stimulus programmes.