Strategic vs Tactical Asset Allocation: The Difference

Strategic asset allocation sets the mix a portfolio holds when nobody has a view. Tactical asset allocation is a bounded, temporary departure from that mix, taken because somebody does. One decides where the portfolio sits by default; the other moves it inside limits the first decision already wrote down. Every other difference between the two follows from that one.

How each decision is carried out is set out separately, under strategic asset allocation and under tactical asset allocation.

The working example throughout is the Anantara Multi-Asset Portfolio, an invented discretionary mandate of Rs 500 crore run for an invented charitable endowment. Rukmini Deshpande chairs the endowment's investment committee and Faiz Ahmad Ansari runs the mandate.

What is strategic asset allocation on this mandate?

Strategic asset allocationThe long term split of a portfolio across asset classes, decided in advance of any market view and written into the holder's own governing document. is the decision about where the portfolio sits across its asset classes when nobody in the room is claiming to know what happens next. The decision is taken once, in advance, and written down.

On the Anantara Multi-Asset Portfolio that decision produced equity 60.0 per cent, fixed income 30.0 per cent and cash 10.0 per cent: Rs 300 crore, Rs 150 crore and Rs 50 crore on Rs 500 crore. The three weights form the policy mixThe stated long term weights a portfolio is designed to hold. They are an intention written in a document, not a measurement of what the portfolio holds today., the holdings the portfolio is designed to have rather than the holdings it has this morning.

The mix did not arrive as a preference. The mix fell out of the holder's own written assumptions, labelled as assumptions rather than forecasts: equity 12.0 per cent expected return at 18.0 per cent volatility, fixed income 7.5 at 5.0, cash 6.0 at 0.5, a correlation of 0.20 between equity and fixed income, and cash uncorrelated with either. Feed those the weights 60, 30 and 10 and the expected return is 0.60 times 12.0 plus 0.30 times 7.5 plus 0.10 times 6.0. The three products are 7.20, 2.25 and 0.60, and they sum to 10.05 per cent. The volatility takes the correlation into account and comes out at 11.20 per cent. The strategic decision is not a number somebody liked, it is the answer a stated set of assumptions gives once the objective and the constraints are fixed.

Notice what the strategic decision does not contain: no statement about what markets will do, no signal, no trigger, no date on which anybody expects to act. The absence is the point. A stated mix is a standing arrangement, revisited on a schedule rather than on a feeling.

What is tactical asset allocation on this mandate?

Tactical asset allocationA deliberate and temporary move away from the stated policy weights, taken because somebody holds a view, and bounded by limits agreed before that view existed. is a deliberate, temporary and bounded move away from the stated policy weights, taken because somebody has formed a view about what happens next. Each of those four words carries weight. Deliberate means decided rather than suffered. Temporary means expected to end. Bounded means a limit written earlier says how far it may go. A move away means it is measured from somewhere.

The somewhere is the policy mix defined above, and the limit is in the mandate the endowment signed: equity may sit between 50 and 70 per cent. Against a policy weight of 60.0 per cent that is ten points of room each way. On Rs 500 crore the edges sit at Rs 250 crore and Rs 350 crore around a middle of Rs 300 crore. The size of any permitted departure was fixed by the strategic side of the arrangement, in a document, before anybody had a view to act on.

A deviationThe difference between what the portfolio currently holds in a class and the policy weight for that class, measured in percentage points. is measured in percentage points from the policy weight, and the mandate bandThe range of weights the holder's own governing document permits for a class. Anything outside it is a breach rather than a position. turns a view into a permitted size.

A band is not a position. The mandate fixes how far equity may move, and until somebody forms a view and acts there is no tactical weight and no signal at all, so every departure below is a constructed one.

What is the single difference the two rest on?

The comparison begins with one question and not with seven. Does anybody hold a view about what happens next? Strategic asset allocation is the answer to where the portfolio sits when the answer is no, and tactical asset allocation is what becomes available when the answer is yes.

A comparison that opens anywhere else starts one step too late. Opening with frequency establishes only that one changes rarely and the other more often. Frequency alone cannot say why the rare decision is still in force on the days the frequent one does nothing. Opening with the view settles that at once: a decision taken without a view has no reason to change on news, and one taken because of a view has no reason to survive it.

The same shape appears without any market in it. A household earning Rs 80,000/- a month puts Rs 20,000/- into a recurring deposit, never to be cut below Rs 12,000/-: a standing decision at 25.0 per cent of income with a floor at 15.0 per cent. With a wedding three months away it moves Rs 15,000/-, or 18.75 per cent, for those three months. The reduction is a departure with an end date, and Rs 20,000/- is what it returns to.

Who takes each decision, and on whose authority?

The two are not taken by the same person. The strategic decision belongs to the holder. Here the holder is the endowment's investment committee, chaired by Rukmini Deshpande, and it decides the purpose of the portfolio and where the portfolio sits by default. The person running the money cannot change it. He inherited it.

The tactical decision belongs to Faiz Ahmad Ansari, and only because the committee handed it over in advance and in writing. A discretionary mandate is that handover: delegated authorityPermission to take a decision on somebody else's behalf, granted in advance and bounded by limits the grantor wrote. The grantor keeps the right to withdraw it., bounded by limits the grantor set and revocable by the grantor. The two decisions sit at different levels of the same arrangement, and only the lower one is delegated.

So the phrase the manager changed the allocation deserves a second question: changed it from what to what, under whose signature? A move from 60 to 66 per cent equity is one person using permission inside a range; a move of the policy mix itself from 60 to 66 is the committee rewriting the arrangement. The portfolio can hold the same securities either way and the two are still not the same act.

Where is each decision written down?

The strategic decision lives in the holder's governing document, beside the objective and the constraints, written as a standing condition rather than as an event: not an act performed on a date, but a standing description of the portfolio itself.

The tactical decision lives in a dated note recording an event: the view, the sleeve that funds it, the size in points and in rupees, the expected life, and what would tell the holder the view was wrong. The strategic record describes a condition and the tactical record describes an act. One therefore has no date on it and the other is useless without one.

The difference between the two records is why the distinction is worth keeping. If the only surviving record of a portfolio is the governing document, what it holds by default can still be stated. If it is a stack of dated position notes, nothing can be stated at all. Each note is measured from a policy weight that is now missing.

What horizon does each one run on?

The strategic mix runs on years. The mix is reviewed rather than traded, on a schedule the holder set, and the review asks whether the objective or the constraints have changed rather than whether the last twelve months were pleasant. A mix reviewed every time somebody feels uneasy is being traded under another name.

A tactical position runs on whatever life the view has, stated in advance along with the test that ends it. A position with no expected life written down has quietly turned into policy, without anybody amending the document that states the policy. The distinction breaks in practice exactly there: not by anybody arguing for it, but by nobody closing anything.

What is each one judged against?

The same number gets pointed at both, and that is where the two are confused most often. The strategic mix is judged on whether it still fits the objective and the constraints, a question about the holder rather than about markets: has the spending need changed, has the tolerance for a fall changed? A year of results is nearly silent on both.

The tactical position is judged against the view that produced it. Did the thing the view described happen, was the falsifying evidence written down beforehand, and did the position close when it should? All three questions are answerable one position at a time.

Put the mandate's own figures against that. The mix was designed to produce 10.05 per cent; the one stated year delivered 14.2 per cent against 12.6 per cent for the composite benchmark, beside a risk-free rate of 6.5 per cent. One year is one draw from assumptions that were never a forecast, and an average is not a promise about any twelve months, so none of that says whether the mix was the right mix.

What does each one assume about the person deciding?

The quietest difference is also the largest. The strategic decision assumes something modest: that the holder can state an objective, the constraints it must sit inside, and a set of assumptions labelled as assumptions. Nothing there requires anybody to know what happens next, only a committee that can describe itself honestly.

The tactical decision assumes three things stacked: that somebody can form a view about what happens next, that the view beats doing nothing, and that it does so often enough to pay for the trading it causes. The third is the one people skip. Every move costs, and the cost does not wait to see whether the view was right, so a view right slightly more often than not, acted on frequently, can still leave the portfolio behind.

The three assumptions are stated plainly. Whether they hold for any person, committee or process is a claim about skill, and skill is settled by a long record of decisions rather than by a stated assumption.

What happens to each if nobody acts at all?

Send everybody away for a year: nobody meets, nobody trades, nobody forms a view. Now ask what the portfolio has.

The stated mix is still the stated mix. A stated mix was never an action in the first place: it is the default positionWhere a portfolio sits when nobody does anything. It holds by standing arrangement rather than by being chosen again each day., the condition that holds when nothing is done. The actual weights did change: prices moved, and the weights moved with them, with nobody deciding it.

There is no tactical position at all. Nothing lapsed and nothing closed: nothing ever happened. A departure is an interventionSomething that only exists because somebody did it. Absent the act there is nothing, as against a default that holds by itself., and an intervention nobody makes is simply absent. One of the two is what remains when nothing is done and the other is only ever what somebody did. Two things that different cannot be compared as rival approaches.

Weights are ratios, so they move whenever any sleeve's value moves, and how far they drift depends on what prices did rather than on anything the mandate fixed in advance.

How is each one brought to an end?

Reversal mirrors authority. A stated mix ends by amendment: the committee finds that the objective or the constraints or the assumptions have changed, and new weights go into the document. From that moment those are what the portfolio holds when nobody has a view. Trades follow, but the amendment is not itself a trade.

A departure ends by being closed back to the stated weight. Whoever runs the mandate buys or sells the amount needed to return equity to 60.0 per cent, and the position is over, with the document reading exactly as it did before. One ends by rewriting the default itself, and the other by going back to a default that never moved.

The two endings give a test for any completed episode: read what the portfolio holds against what the document states, and one of three answers follows. The 66.0 per cent used below is a constructed weight, since no mandate fixes a departed weight in advance.

How much of the portfolio does each decision actually settle?

The sizes follow from the one assumption set stated at the top. The strategic decision put the whole result on the table: at 60, 30 and 10 the expected return is 10.05 per cent and the volatility is 11.20 per cent, not the increment from anything.

Now take the largest departures the mandate permits, holding fixed income at 30.0 per cent so that one thing moves at a time. At the lower edge, equity at 50.0 per cent with cash absorbing the difference at 20.0 per cent, the figures are 9.45 and 9.42 per cent. At the upper edge, equity at 70.0 per cent funded out of cash so cash falls to zero, they are 10.65 and 12.98 per cent.

Subtract from the stated point. The most a departure can add to expected return is 10.65 less 10.05, or 0.60 percentage points, and the most it can remove is the same 0.60. On volatility the same subtraction gives 1.78 points either way. Against a designed 10.05 and 11.20 per cent, those are 5.97 per cent and 15.89 per cent of themselves.

The tactical decision is a small adjustment to a large one, and treating the two as competing approaches has the relative sizes wrong by an order of magnitude.

Walk the equity weight across the permitted range in steps of two points, letting cash absorb each step, and the eleven computed pairs trace a short arc. Anywhere on it is admissible; nowhere off it is.

The tactical decision chooses a point on that arc; the strategic decision chose where the arc is, between 9.42 and 12.98 on an assumption set that runs from a class at 0.5 per cent volatility to one at 18.0.

One more computed result belongs here, and no departure can produce it. The three volatilities in the stated proportions come to 0.60 times 18.0 plus 0.30 times 5.0 plus 0.10 times 0.5, or 12.35 per cent. The portfolio volatility would be exactly that if the sleeves always moved together. The computed 11.20 per cent sits 1.15 points below that, and the gap is the diversification. The gap exists only because the correlation is 0.20.

Every weight used above is then checked against the band the holder wrote. The eleven run from Rs 250 crore to Rs 350 crore, all inside it, the two end points exactly on the edges.

How do the two compare, criterion by criterion?

Read the summary downward rather than across. The rows are ordered so that the first one causes all the others.

Can a portfolio run one of them without the other?

Yes in one direction and no in the other. A portfolio can hold a stated mix, review it on a schedule, restore its weights when they drift, and never take a departure in its life. Such a portfolio is complete: every rupee has a stated reason for being where it is.

A departure needs something to depart from, so no portfolio can run a departure process with no stated mix. Overweight equity, relative to what? Sized from where? Closed back to which weight? Every sentence a tactical process needs to speak contains a reference point only the strategic decision can supply, so one is a prerequisite for the other rather than an alternative to it.

What does the record refuse to supply?

A clean two column layout wants a number in every cell, and that is exactly when a comparison invites invention. Where the mandate record is silent, the honest entry is the words not supplied rather than a plausible number.

How does a practitioner use this distinction in a room?

Faiz Ahmad Ansari uses it as a filter on his own reporting. Everything reaching the committee is sorted into two piles: this is the arrangement working by standing agreement, or this is something I did because I held a view. The first needs no defence; the second needs a note with a view, a size, a funding source and an exit test.

Rukmini Deshpande, chairing, uses it to protect the agenda. The stated mix is reviewed at its own scheduled meeting and nowhere else, so a discussion of last quarter cannot quietly turn into a rewrite of the portfolio's own purpose.

From outside a mandate the same idea becomes a two question filter needing nothing confidential, and a household can ask the same pair about a standing monthly transfer and any temporary change to it.

When does the distinction stop mattering?

Four conditions make the two decisions produce the same portfolio, and on a mandate carrying any of them the distinction is worth understanding once and then setting down. Three of the four can be read straight off a document.

The first is a band too narrow to move anything. Suppose the endowment had written 58 to 62 per cent for equity instead of 50 to 70. Two points of equity is Rs 10 crore on Rs 500 crore, against Rs 50 crore at the wider band. Expected return then runs between 9.93 and 10.17 per cent against a designed 10.05, about 1.2 per cent of it either way, and volatility between 10.84 and 11.55 against 11.20. Whether the manager sits at an edge or at the centre, the holder is looking at the same portfolio to within a tenth of a point.

The second is a mandate that grants no band at all. Where the document states weights and no room around them, no tactical decision is available to anybody and every question in this guide collapses into the strategic column. A great deal of money runs that way.

The third is a horizon that outruns the position. A departure with a stated life of three months, closed back to 60.0 per cent, moved the portfolio for one quarter of a horizon measured in decades, and the weight it left is the weight it returned to.

The fourth is the ordinary day, on which the view agrees with the policy weight. Whenever the evidence says 60.0 per cent is where equity belongs, both answers are the same number and the portfolio holds Rs 300 crore of equity either way.

None of the four announces itself when it ends. A band gets widened at a review nobody called a change of approach. A discretionary permission is added on renewal. A horizon shortens because the endowment takes on a spending commitment. On the day any of those happens the distinction is load bearing again and nothing arrives to say so. The distinction is therefore worth being able to state even where it currently decides nothing.

Whether running a tactical process is worth it for any holder turns on whether the three stacked assumptions hold for a particular committee. Skill is the claim being made, and one year of figures cannot settle it.

The Anantara Multi-Asset Portfolio, the charitable endowment that holds it, Rukmini Deshpande and Faiz Ahmad Ansari are invented.
Educational material. Not advice on any investment, tax, budget or market position.