974 words · 4 min read
Glacier CapitalEssays
Home/Essays/VOLUME VIII · A NEW CHAPTER/NO. 7
VOLUME VIII · A NEW CHAPTER · TALENT · NO. 100 · CONTRIBUTION ONLY

Not Where You Came From, Only What You Contributed

No questions about degree, title or years in the business — only whether you exceeded expectations on this mandate.

Zhang Jiakang (JK) · Founding Partner, Glacier Capitalapprox. 676 words · 3 min readArchived 2026-08-16

A colleague who had recently joined once asked us: am I doing well, have I exceeded expectations?

The question is heavier than it sounds. What he wanted was not praise but a ruler. Let me state the view up front: for the second eight years, our standard for assessing people converges into one line — not where you came from, only what you contributed. Why? Three layers below.

Not Where You Came From Means Three Questions We Do Not Ask

Not the degree, not the title, not the years in the business. However much you did is however much you take.

In the Archive this comes down to two things. First, we follow and accompany people from the doctoral stage at top universities, building the early talent pipeline inside the schools; in other words, this train does not leave only in hiring season. Second, new members take a deep part in core transactions in their first year, with the mature methodology settled in the organisation underneath them. The first is about developing people; this piece is about measuring them. Development can be patient. Measurement has to be clean.

Why Not Allocate by Rank?

It does not work. The reason is not values. It is arithmetic.

Allocating by rank assumes rank represents contribution accurately. In project work that often fails to hold: sixty days and eight checkpoints a mandate, and an associate may have led half of them. Someone will say, then make the rules finer and account for every unit of contribution. But this business sells judgement, and judgement cannot be attributed cleanly. Contribution on a banking mandate is a web, and the chain of tracing only gets longer. Rules that mean to abolish ambiguity end up turning ambiguity into dispute. The more you calculate, the less clear it gets.

So we keep one thing: rank can be broken through. There is only one basis for breaking through it — whether he genuinely exceeded expectations on this mandate. That is hard to quantify. But a distorted allocation table hurts more than ambiguity does.

The Other Side of Counting Contribution Is Not Counting Toil

Hours present are not output. Exhaustion should not be honoured. We wrote health and rest into the organisation's rules, and that is not a benefits clause but protection for the quality of judgement. This business sells judgement; across twenty to thirty mandates a year, each one needs someone clear-headed on the few days that matter. Promotion looks only at whether a benchmark project actually landed. Overtime hours? Not counted at all.

The risk here is real. Hiring with a developmental mindset in a harsh environment is itself a high-risk move: when the market is tight, developing people takes up exactly the scarcest front-line time — the most expensive supplies of the year. A boutique investment bank is not a training school. That sounds cold, but it is honest to the person being hired: your first year will not raise you as a student, it will use you as a colleague.

So what makes us willing to use new people this way? The organisation underneath them. Method gives the path, context gives the raw material for judgement, and only then dare an individual be measured cleanly. A clean ruler is why young people are willing to come. This is not softness; the arithmetic works: sitting at the table of real transactions in your first years gives the learning curve a completely different slope. Talent density sets the ceiling of an organisation, and team culture decides how long that ceiling holds — the first is instantaneous compute, the second is long-term bandwidth.

On this we are fairly confident: an institution can be a philosophy rather than one person's company. If the glue is one person, the structure comes apart when that person leaves; if it is a shared way of judging, it can cross generations. Will some people not take to this ruler? Some will not.

If you would like to be measured by it too, the talent contact address is in the Archive: next@glacierchina.com. The rest, in person.

The Archive entry for this piece