Rebuilding the bond side: from overnight cash to a ladder, credit and index-linked

1. The situation

A UK family portfolio, long-horizon, currently being drawn on at a low rate — under 3% a year. The equity side is deliberate and I am not touching it: a global market-cap core, a value-factor sleeve, an ex-US sleeve, and a small commodity/mining tail. The intent there is unchanged and stated in my own notes: own the whole market cheaply, accept that the US is ~60% of it, and take the risk in equities rather than in the ballast.

The problem is what had accumulated on the bond and cash side. Over a couple of years of moving in and out of things, it had drifted into a pile of very short instruments — money market, ultrashort corporate, and gilts maturing inside five years — spread across three wrappers. Nothing had been chosen. When I finally counted it up:

• the “spending ladder” (cash plus ultrashort plus 0–5-year gilts) covered nine years of drawdown;
• credit: zero. Index-linked: zero. Anything maturing beyond the early 2030s: zero.

So the income side of the portfolio was doing the one thing that is almost guaranteed to underperform over a decade: holding everything overnight, for years it will not be spent for, and calling that safety.

2. What I am changing, in three moves

a) Keep the equity position as it is. The mix of the equity sleeve is unchanged. This is a bond-side project; the growth engine is not the subject of it.

b) Reduce gold from 10% to 7.5%. Gold is insurance, not an investment: it pays nothing, it has no earnings to grow, and its long-run real return is roughly zero. I still want the insurance — the proceeds are not leaving the portfolio — but 10% is a large premium for a hedge whose payoff is undated, so it goes to 7.5% and the money goes into the bond side.

c) Rebuild the bond side into four distinct jobs. This is the substance:

Job	Rough share of the bond side	What it does
Ladder (cash float + ultrashort + 0–5yr gilts)	~40%	money I expect to spend in the next few years; barely moves when rates change
Sterling credit (investment-grade corporate)	~25%	the yield: lending to companies rather than only to the government
Index-linked gilts	~18%	defends buying power if inflation surprises upward
Global aggregate, GBP-hedged + longer gilts	~17%	diversification away from lending only to one government

2b. The change in one picture

Percentages of the bond-and-cash side only. Nothing here is an amount.

pie showData
    title Where the bond side was
    "Ladder: cash, ultrashort, 0-5yr gilts" : 76.5
    "Index-linked" : 12.1
    "Duration: one gilt to 2030" : 11.3

pie showData
    title Where it is going
    "Ladder, trimmed to a 3-year floor" : 32.6
    "Sterling investment-grade credit" : 25.1
    "Index-linked gilts" : 18.0
    "Global aggregate (hedged) + longer gilts" : 24.3

Two things to read from the pair. First, the ladder falls from roughly three quarters of the bond side to about a third — that is the whole change. Second, the credit sleeve goes from nothing to a quarter, which is the only place where the portfolio is genuinely taking new risk.

3. The bit I would most like challenged: the ladder

I am cutting the ladder from nine years to a floor of three, and holding the second and third years in ultrashort credit and short gilts rather than in overnight money. The reasoning:

• the draw is under 3% a year, funded first by dividends and coupons, so forced selling is unlikely;
• ultrashort and short-dated funds are liquid within a day or two and largely investment-grade, so the liquidity barely changes;
• overnight money is the lowest-yielding rung on the ladder, and holding nine years of spending in it was paying for safety I do not need.

The cost is real and I want to name it: I am taking a small amount of credit and duration risk in the money I might spend within a few years. If a recession arrives, credit spreads widen and the second and third rungs of the ladder fall together — that is the trade.

A separate argument inside my own policy is whether the ladder floor should be three years or five. The bond-side planning says three is sufficient given the low draw; the older rule of thumb says five to seven. I have changed the rule, and I would like to know whether that is a genuine improvement or just a rationalisation of wanting a better yield.

4. The instruments

All sterling-exposure — deliberately, so the ballast cannot be knocked over by a currency:

• short-dated gilts (0–5 years) for the back of the ladder;
• ultrashort investment-grade — supranational and corporate paper maturing inside a year;
• sterling investment-grade corporate bonds for the credit sleeve — the cheapest tracker available, roughly 0.20% a year, several hundred holdings;
• UK index-linked gilts — the fund holds around 35 index-linked gilts, essentially all UK;
• global aggregate bonds, hedged to sterling — about 0.08% a year for the whole global investment-grade market with the currency risk removed;
• UK gilts of all maturities for the duration sleeve — around 0.05% a year.

Blended cost across the whole bond side lands near 0.11% a year, against 0.40–0.60% for the equivalent “core-plus” mandates — but the point of doing it this way was visibility, not 30 basis points. I want to be able to see what each sleeve is doing and why.

5. What I am deliberately not doing

• No duration bet. No 20-year+ Treasuries or 15-year+ gilts as a rates view.
• No unhedged foreign bonds. A currency position inside the ballast is not what the ballast is for.
• No high yield or EM debt at this size. The extra spread is real, but so is the correlation with equities exactly when I need the ballast to behave.
• No tactical timing. The change is driven by what the portfolio lacked, not by a view on the next 12 months.
• No all-cash safety. That was the position I am unwinding.

6. The five questions I would put to a former bond dealer

1. Ladder floor: three years of drawdown in short/ultrashort — sensible, or should it be five?
2. Credit spread: at current sterling IG spreads, is the move out of gilts into corporate credit being paid enough, or is the market already pricing a benign outcome?
3. Index-linked at these real yields: is ~18% of the bond side a reasonable inflation hedge, or is that a lot of money in an instrument that has had a brutal decade?
4. Hedged global aggregate vs gilts: is the global sleeve worth its complexity once hedging costs are counted, or are UK gilts plus credit enough?
5. Sequence risk: moving from overnight cash to ultrashort credit in the sleeves I may spend from within three years — is that a mistake if the next recession arrives early in a drawdown phase?

7. What would change my mind

Stated in advance, so this is a decision and not a mood:

• If sterling IG spreads are tight against their own history, the credit sleeve gets built more slowly and the ladder stays longer.
• If the draw rate rises materially, the ladder floor goes up, not down.
• If over the next two years the ultrashort sleeve does not beat the cash it replaced after its own costs, the honest conclusion is that the ladder should be cash and the risk belongs in equities — and I will say so.