Ask why a Margaux smells the way it does and you will usually be told it is the terroir, which is true in the way that "because of physics" is true. It closes the question rather than opening it. The useful version takes about four sentences, and once you have it you can predict roughly how any Medoc wine will behave before you pull the cork.
Start with the stones. All of the Medoc's good vineyard land sits on banks of gravel that the rivers dumped there over a few hundred thousand years, sitting on top of clay and limestone. The gravel is the point: it holds almost no water and no nutrition, it warms up in the sun, and it forces the vine's roots down through it to find whatever is underneath.
Why the south is perfumed and the north is firm
The banks are not the same everywhere. In the north, around Pauillac and Saint-Estephe, they are deep and they sit over more clay. In Margaux, at the southern end, the gravel tends to be shallower and the ground drains harder. Three things follow from that, and they are the whole explanation.
First, the vines run short of water earlier in the season. A mildly thirsty vine stops pushing out leaves and shoots and puts what it has into the fruit, and it makes smaller berries. Second, shallow stony ground over good drainage warms faster, so Margaux ripens earlier than the northern communes and is usually picked first. Third, ripening earlier means picking at a point where the aromatic compounds in the skins are at their most floral and before the tannins have built the mass they build further north.
That is the perfume. It is not mystical. It is a warmer, poorer, faster-draining piece of ground producing smaller berries picked a week or two sooner, which gives you violet and dark red fruit on a medium frame instead of graphite and cassis on a heavy one.
Here is the complication. Margaux is the largest of the four famous communes and by some distance the most mixed. The appellation covers five villages, and the ground under them ranges from textbook gravel mounds to patches that are frankly clay. "Margaux" on a label is therefore the least reliable of the four commune names as a prediction. You still have to know the property.
Dauzac
Dauzac sits at Labarde, at the southern edge of the appellation, on flat ground close to the river rather than up on the high gravel mounds. It is a fifth growth, and for a long time it was the name people reached for when they wanted an example of a classified property coasting.
The estate also has a footnote in agricultural history. In the 1880s, the story goes, its manager was painting the roadside vines with a mix of copper sulphate and lime to stop passers-by helping themselves, and noticed those vines were not getting mildew. That mixture became the Bordeaux mixture, still sprayed on vines and vegetables worldwide, including on a good deal of what is sold as organic.
What matters more for anyone buying a bottle is that Dauzac has changed hands and changed direction repeatedly over the last forty years: institutional owners, consultant winemakers, new cellars, new teams. This is where the standard way of thinking about Bordeaux quietly breaks. A chateau is treated as a fixed thing, a place with a character. Roughly half of what is in the bottle is the ground, which does not move, and the other half is a set of decisions about when to pick, how hard to select and how much new oak to use. Those decisions belong to whoever is currently employed there.
So a bottle from 1981 and a bottle from 2013 are not two data points about the same wine. They are two different wines made from the same field.
What we have
Our Dauzac range is a ladder with a long stretch missing in the middle, which is normal for a property whose old vintages come to us in ones and twos.
The bottom rung is the modern estate: 2013 at around HK$370, in real quantity, the only bottle here you can buy a case of. Then a jump to the library, where the bottles are older than most of the people who will drink them. 1981 at around HK$780, 1997 at around HK$890, and a single 1985 at around HK$1,180. At the top, a six-litre 1997 at around HK$4,400, which is a wine for a room rather than a dinner.
Climbing this ladder does not get you a better-made wine at each step. It gets you an older one, in shorter supply.
What we would pick, and why
2013, around HK$370. 2013 was a genuinely difficult year in Bordeaux, and the market has priced it accordingly ever since. What that produces is a light, early-drinking claret with modest alcohol and no need for a cellar, from an estate that had every incentive to select hard because there was nothing to be gained by pushing volume in a bad year. For a Tuesday bottle with roast pork or a plate of char siu, low alcohol and a light frame are advantages, not apologies. This is the pick.
1981, around HK$780. Forty-five years old, for less than a thousand dollars. 1981 is not a famous vintage, which is the only reason the number is what it is. A mature Margaux at this price is a bottle you can afford to be wrong about, and being wrong is a real possibility with anything this old. Open it with people who will find a tired bottle funny rather than disappointing.
The six-litre 1997, around HK$4,400. Not because the wine is better. Large formats age more slowly per litre and a six-litre bottle will hold a table of twenty. If you are hosting a birthday and want one object on the table instead of eight, this does that job and no 75cl bottle does.
Where paying more stops buying flavour
Somewhere above the 1981, the money stops going into the glass. The 1985 at around HK$1,180 costs half again as much as the 1981, and the honest reason is that 1985 is a better-regarded year and there is one bottle of it. That is scarcity plus reputation, and both are real things that cost real money, but neither of them is flavour. If your budget is fixed and you want the most wine per dollar out of this page, you should be buying the 2013 by the case and stopping there.
Two more warnings. Anyone buying the older bottles as a store of value should not: these are drinking bottles from a fifth growth, not a market. And anyone buying 2013 expecting the thick, dark, high-alcohol style that a lot of Hong Kong palates were trained on in the 2009 and 2010 vintages will find it thin. It is not thin. It is a cool year, honestly made and honestly priced, and it wants food to lean on.