Whoever knows the businessman can know the risk before the bank does.

A financial statement shows what the company declares. A long economic relationship shows how it behaves.

I have spent the last few years trying to understand why some people see a credit operation where others see merely a company asking for money. For a while I thought the difference lay mainly in access to capital. Then I saw that capital, at certain moments, is precisely what is in surplus. The hard part is finding something good enough to receive it. That led me to origination, to the quality of deal flow, and to the importance of knowing how to turn a business need into an operation that makes sense for a given creditor.

Now I am beginning to suspect there is an advantage prior to all those things: knowing the borrower before he becomes a borrower.

Amazon announced this week that it has already lent more than US$ 3 billion to small companies selling on its platform. More than US$ 1 billion in the last twelve months alone. More than twenty thousand companies have already received some financing. The program began in 2011 and offers short-term loans, up to twelve months, to sellers invited by Amazon itself.

The word that interests me most is invited.

The merchant does not simply walk into a branch and try to convince Amazon that he has a good business. In many cases, Amazon has been observing that business for quite a while. It knows how much he sells on the platform. It knows how often he sells. It knows part of the history. It sees when sales grow, when they fall, how much inventory turns, and how the company behaves inside that ecosystem.

The credit arrives after the economic relationship.

That inverts an important part of the traditional process.

Normally, the financial institution meets the businessman at the moment he needs money. It is a curious circumstance in which to begin a relationship. The man walks into the room precisely when he has an incentive to present the best possible version of his reality.

Naturally he hands over balance sheets, statements, bank records, financial information, and collateral. The bank has techniques, models, and professionals to evaluate that. There is nothing wrong with it. The problem is that much of the analysis is done looking at records produced before the relationship between the two parties existed.

Amazon occupies another position.

It observes before lending.

That can be worth a lot of money.

Not because data eliminates risk. That would be a conclusion typical of someone who discovers a new source of information and immediately declares everything that existed before dead. Financial history is generous with people who believed they had finally eliminated the need for judgment.

Data does not eliminate risk.

It changes what we can see.

A company can present a correct balance sheet and still hide bad behavior. It can hold assets and treat suppliers as a permanent source of involuntary financing. It can be profitable and manage cash imprudently. It can pay every obligation, but only after several collection attempts.

Those things leave traces.

Not always in the same document.

Sometimes in the relationship.

A supplier who has sold for fifteen years to a given company may have an understanding of its risk that no score can reproduce instantly. He knows whether the owner asks for extra time when the market tightens. He knows whether he honors what he promises. He knows whether he warns before being late or disappears until chased. He knows whether the purchases are increasing coherently with the growth or whether something is odd.

The accountant knows another part.

The lawyer knows another.

The long-serving employee may know another.

The bank knows what passes through the account.

None of those agents holds the whole truth.

But some hold observations that do not appear when a new creditor looks only at the photograph.

That may be one of the great limitations of financial models: they have to turn behavior into measurable information after the behavior happened.

The relationship observes the process while it happens.

It is a subtle difference.

A financial statement can report that a company sold R$ 100 million last year. A commercial partner may be able to say whether those R$ 100 million were produced healthily, whether the businessman destroyed margin to grow, whether he lost important clients, whether he started granting absurd terms, or whether he is merely crossing a temporary moment.

That does not make the partner a better analyst than the bank.

It makes his information different.

Perhaps good credit appears when we can combine the two.

I am beginning to think the greatest informational advantage is not necessarily in holding a secret piece of data. It may be in observing the same man in different contexts for long enough.

Reputation is a kind of human time series.

It should not replace numbers. But it should not be ignored either merely because it does not fit into a spreadsheet.

That interests me particularly because it may change the role I imagine for a loan broker.

Until recently, I thought of the broker above all as someone able to identify the need, structure an operation, and know sources of capital diverse enough to find compatibility. I still believe that. But I see an even more valuable function: carrying context along with the operation.

A bank can receive fifty companies in a week.

For the analyst, all of them start as files.

For an originator who has known a certain company for years, one of them starts as a story.

The question is turning a story into useful information without turning proximity into blindness.

That distinction will be hard.

The relationship creates an advantage and a threat at the same time.

Knowing the businessman lets you see behavior.

Liking the businessman can make you ignore behavior.

They are not the same thing.

There is a human tendency to treat trust as an objective reduction of risk. Sometimes it is. A man who honored commitments for twenty years has a relevant track record.

But trust can also be merely repeated familiarity.

I know him, therefore he seems safe.

It is precisely that association that has to be fought.

The good originator should use the relationship to formulate better questions, not to dispense with them.

Perhaps his advantage is knowing where to ask.

He knows the moment the company grew too fast. He knows there was a shareholder dispute two years ago. He remembers that a certain client represents a large part of the revenue. He knows of a property that could improve the collateral. He knows the businessman does not want to dilute his stake.

That information does not approve the credit.

It helps build the analysis.

That is a far more interesting role than simply sending documents.

Perhaps the loan broker of the future has to function partly as a layer of intelligence between the company's real life and the financial system.

On one side is the businessman, whose reality is disorganized, continuous, and full of particularities.

On the other, an institution that has to turn that into policy, numbers, limits, and a decision.

Someone has to translate.

The better the translation, the smaller the chance of a good risk being refused out of incomprehension and a bad risk being approved because of a well-made presentation.

That is a form of information arbitrage that interests me.

Not arbitrage in the sense of holding privileged or improper information. That would be another subject. I mean seeing something the standardized process sees badly.

Large systems have to standardize because scale requires standardization.

Standardization produces blind spots.

It is almost inevitable.

A bank cannot organize its policy around every peculiarity of every businessman. It has to create boxes.

Minimum revenue.

Maximum indebtedness.

Sector.

Tenor.

Accepted collateral.

Rating.

Score.

Documentation.

That makes the system manageable.

It also leaves good deals between the boxes.

That is where private capital can appear.

Not to finance what the bank correctly considered bad.

That would be the fastest way to turn private credit into a dump for rejects.

Opportunity exists when institutional refusal and economic risk are not exactly the same thing.

Perhaps the bank does not accept a certain piece of collateral because operationally it does not work with it.

Perhaps the tenor does not fit the product.

Perhaps the ticket is too small for a fund and too large for a given banking line.

Perhaps the businessman has good cash flow, but the company is going through a reorganization that makes the historical numbers unrepresentative.

Those cases require context.

That is where off-market credit starts to look especially interesting.

Not because the private market is magically smarter. In many cases it will be less transparent, more expensive, and more subject to conflicts.

The advantage lies in the ability to build a specific operation for a specific situation.

A family office, for example, can look at a given credit in a way a much larger institution cannot.

Not because it has better analysts.

It may have something more important in that case: flexibility.

The family may know the sector.

It may know the originator.

It may understand the property offered as collateral particularly well.

It may have a five-year horizon when the bank needs two.

In that case, the structure can make sense for both.

That helps explain why some families start investing directly in companies and private operations.

What they are looking for is not merely a higher return.

It is some form of advantage.

Relationship.

Knowledge.

Control.

Access.

The possibility of structuring.

That market seems more and more relevant among family offices. This year, the Family Office Exchange itself has been devoting its studies and forums to themes such as investment strategy, allocation, and direct investing among families. It does not appear to be a marginal curiosity. It is a natural evolution for estates that stop wanting to buy only what has already been packaged for broad distribution.

The problem, as always, is that access produces its own risks.

A family office receives an opportunity because it knows another family office.

The relationship may mean there is a good origin.

It does not mean there is a good operation.

That may be the main intellectual test of private markets: managing to use trust to reduce transaction cost without letting it reduce the quality of the analysis.

A club deal among well-acquainted families can be excellent precisely because each brings capital, knowledge, or relationships. It can allow acquiring a company, financing an asset, or taking part in an investment too large for one family alone.

But there is a trap.

When everyone at the table knows each other, nobody wants to ask the offensive question.

That is extremely dangerous.

The offensive questions are usually exactly the ones that preserve money.

Who verified?

Who originated?

How much does the originator earn?

Who is putting up his own capital?

Why is the seller leaving?

What is the downside?

Who controls?

How do we exit?

If the answer to those questions is replaced by "we have known the family for many years," the relationship may be increasing risk instead of reducing it.

The best financial relationship should allow harder questions, not require softer ones.

That may be a good definition of mature trust.

It is not ceasing to verify.

It is being able to verify without the other person treating the verification as an insult.

There is something similar in the credit relationship between businessman and loan broker.

If the client believes that organizing documentation, opening information, showing debts, and explaining problems means a lack of trust, the relationship is badly built.

A serious broker should know the client well enough to say that a certain piece of information has to appear before the bank discovers it on its own.

Nothing destroys trust with a creditor faster than discovering mid-analysis what the originator apparently already knew.

The problem is not merely the information.

It is the sense that there was an attempt to manage perception.

Financial reputation is extremely sensitive to that.

Perhaps that is why the genuinely good broker sometimes has to contradict his own client.

"That debt has to appear."

"That problem has to be explained."

"I will not send it before receiving those documents."

"That collateral is not worth what you believe."

"That number has to be corrected."

The intermediary who merely reproduces the borrower's narrative is not adding intelligence.

He is outsourcing his reputation to the optimism of someone paid for the approval.

That looks like a fragile business.

The more I think about it, the more I believe the originator's true position lies between representation and judgment.

He represents the client before the market.

But he cannot abdicate judgment entirely in the client's favor.

If he does, institutions quickly learn to discount his words.

Perhaps a broker's reputation is a kind of informal rating on the quality of the first analysis.

It does not replace the formal rating.

But it changes the starting point.

It is the difference between the institution thinking "one more company wanting credit" and "worth a look because it came from here."

That small change can be worth a great deal.

A good creditor's attention is scarce.

Capital may not be.

There are billions available in the financial system, but there is limited capacity of good analysts to examine every request. A badly prepared operation consumes several people's time before dying.

Whoever reduces that waste creates value even when he puts up no capital of his own.

That may be one of the most elegant forms of intermediation.

Not owning the capital.

Not taking the whole decision.

But improving the quality of the information that reaches whoever decides.

That position is beginning to look a lot like what I imagined years ago when I asked how someone could sell money without owning the money.

The answer is becoming more precise.

A relationship with the borrower.

Knowledge of the sources.

The ability to translate.

Enough reputation for both sides to listen.

And process, so as not to depend merely on memory and impression.

That last part still bothers me.

The more I value relational information, the more I see the risk of it staying only in someone's head.

Imagine a loan broker with a hundred clients.

He knows one of them had a certain problem three years ago.

Another has an excellent property.

Another prefers a certain structure.

He knows managers at fifteen banks, different policies, funds, private investors.

It is an extraordinary asset.

It is also chaos.

If everything depends on memory, growth begins destroying the very advantage the relationship created.

The professional knows ten clients so well that he can do excellent work.

With fifty, he starts forgetting.

With a hundred, he has to simplify.

At that moment, the relationship risks becoming merely a record again.

Perhaps the challenge lies in turning relational knowledge into institutional memory without reducing people to CRM fields.

That idea interests me especially when I think about family offices.

A good family office should know a family almost the way a businessman knows his own company.

Know where the wealth came from.

Which assets are untouchable for non-economic reasons.

Where concentration exists.

What debts exist.

Which opportunities are of interest.

Which risks the family already runs through the operating business.

Who decides.

Who will inherit.

Which banking relationships exist.

Which investments went wrong and why.

That is far more information than an investment portfolio.

It is context.

Perhaps the true product of a family office is exactly preserving that context over time.

A family should not have to retell its financial life every time a new investment appears.

The institution should already know.

That completely changes the quality of the advice.

A bank looks at a loan.

A manager looks at a portfolio.

A lawyer looks at a legal structure.

An accountant looks at taxation.

A family office, at least in theory, should see the interactions among all of them.

It is the difference between specialists looking at separate organs and someone looking at the organism.

That coordination may look unglamorous next to picking large investments.

It may be more valuable.

An excellent investment in the wrong portfolio can be a bad decision.

An expensive credit that avoids selling an even more valuable asset can be rational.

An extraordinary off-market operation can be inadequate if it increases a concentration that already threatens the family's wealth.

Context changes the meaning of the decision.

That is exactly what Amazon appears to have with regard to the sellers it finances: context.

It does not have to look at the small business as though it were a complete stranger. An economic relationship already exists.

The company knows that a given merchant sells his products through the platform. It sees some part of the real demand. It knows the internal seasonality. It can observe, within the limits of what it holds, signs of growth and behavior.

The difference from a completely new creditor is enormous.

That may be one of the paths of credit in the future.

The institution that has operational data before it has the credit relationship can better assess certain risks.

An acquirer sees receipts.

A commerce platform sees sales.

Financial software can see cash flow.

A payroll company can see wages.

A marketplace sees inventory and demand.

Every economic activity produces data that may one day become underwriting.

That has interesting consequences.

First, the border between a technology company and a financial institution becomes less clear.

Second, a commercial relationship can turn into credit origination.

Third, the bank stops being automatically the organization that knows the client best merely because it holds his account.

That last change may be enormous.

For decades, the bank held a natural advantage: it saw the financial movement.

Now non-bank companies are starting to hold equally useful data, perhaps even richer in certain contexts.

The bank sees that R$ 100,000 came in.

The platform may know which products produced the R$ 100,000.

That does not eliminate the banking advantage.

But it fragments the information.

The future of credit may belong, in part, to whoever can assemble those fragments.

That makes me think again about the originator's position. Perhaps his job is precisely reconstructing a view the system fragmented.

The businessman has accounts at different banks.

Investments at a brokerage.

Properties in separate structures.

Receivables.

Stakes.

Debts.

Documents at different offices.

No single institution sees the whole economic person.

One sees the account.

Another sees the taxes.

Another, the wealth.

Another, the business.

The originator who can organize that information can present the creditor with a reality far closer to what actually exists.

I am not talking about dressing up a balance sheet.

The opposite.

Removing noise.

A businessman can look riskier because his wealth is badly documented.

Another can look less risky because his presentation is impeccable.

The quality of the organization should not be confused with economic quality.

But in the real world, it will be.

Capital decides on the basis of what it can see.

So making reality legible has value.

Perhaps an important part of financial work is simply that.

Legibility.

A disorganized company can have excellent assets.

While nobody can understand them, the market will apply a discount.

A family estate can be enormous and still negotiate badly with banks because it appears fragmented.

If someone can organize all the assets, debts, flows, and collateral into a coherent view, he can change the negotiating position without creating a single new cent of wealth.

That is fascinating.

The money already existed.

The intelligence lay in presenting it as a system.

It may be one of the clearest examples of how information can change price without changing the asset.

A bank can charge an opaque borrower dearly because it has to be paid for uncertainty.

If part of the uncertainty comes from bad organization rather than real economic risk, there is possible arbitrage.

Organize better.

Show better.

Structure better.

The spread can fall.

In other words, information is also collateral, though it does not replace legal collateral.

It reduces the part of the risk born of not knowing.

That may be especially important for businessmen who do not live inside financial language. An excellent industrialist can be a terrible presenter of credit.

He knows his business by the smell of the factory, the rhythm of the orders, the conversations with clients and suppliers. An institution's analyst has never seen the factory. He has to turn all of that into statements, ratios, contracts, and projections.

Someone has to build the epistemological bridge between the two.

The better he does it, the larger the credit market available to the businessman.

That starts to look like a service of great value.

Perhaps the client pays not merely to obtain money, but to become financeable before more sources.

Those are different things.

Obtaining money once solves an operation.

Building a coherent financial representation can reduce the cost of capital for years.

That second service comes closer to a family office than to a traditional broker.

The family office can organize wealth, information, and relationships so that, when a need for capital appears, it is not necessary to rebuild the client from scratch.

The history already exists.

The documentation is already organized.

The collateral is known.

The positions are mapped.

The indebtedness is monitored.

The adviser knows the family.

That drastically reduces the cost of a decision.

That may be one of the reasons very wealthy families have access to operations that people with smaller estates consider "exclusive." It is not merely because the bank likes them more.

There is infrastructure around the wealth.

Lawyers.

Bankers.

Advisers.

A family office.

Relationships of decades.

Information ready.

When an opportunity appears, those people can act.

The small investor has to start by looking for the phone number.

The large estate already has the network before the opportunity.

That is an accumulated advantage.

Perhaps social capital and financial capital compound together.

Money creates relationships.

Relationships create access.

Access creates opportunities.

Good opportunities increase wealth.

Wealth further increases the value of the network.

There is a cycle.

But it is not guaranteed.

The same network can turn into a machine for distributing bad deals if the family starts confusing proximity with quality.

That is why I want to keep insisting on the filter.

A relationship should increase information.

Never replace judgment.

That may be the rule of 2017.

Knowing someone can make you understand the risk better.

It should not let you ignore the risk.

That holds for borrowers.

It holds for investments.

It holds for club deals.

It holds for other family offices.

It holds for any market where reputation and money circulate together.

There is something especially dangerous about operations between socially close people because financial enforcement can be read as a breach of trust. That is why contracts, structures, and independent agents may be even more necessary, not less.

If five families make a private credit together, the prior relationship makes the origination easier.

After that, the operation should be treated professionally.

Documents.

Collateral.

An agent.

Governance.

Periodic information.

A plan for default.

The families should stay friends even if the operation goes wrong, which requires the rules to have been set while everyone was still happy.

Good governance preserves relationships because it removes from the friendship the obligation to resolve what should have been resolved by the contract.

The same holds for a broker financing a close client.

It may be necessary to be even more rigorous.

The friendship has already introduced an emotional variable.

The structure has to compensate.

I have more and more sympathy for mechanisms that admit our weaknesses instead of presuming that professionals suddenly become immune to them.

We are all capable of rationalizing.

The system should make some rationalizations harder.

An independent committee.

A second opinion.

Exposure limits.

A checklist.

Mandatory documentation.

They may look like bureaucracy when the professional is brilliant.

It is exactly when someone considers himself brilliant that they may be most necessary.

Overconfidence is a risk that grows alongside a positive track record.

The more often we are right, the less able we are to imagine that the next success might depend partly on luck.

That is a reason to like long historical data and, simultaneously, to distrust it.

The track record helps.

But the environment changes.

A company that paid correctly for ten years can enter a new business, change management, or take on too much debt.

A relationship does not freeze risk.

The advantage lies in noticing the change before the annual document.

Perhaps an old supplier notices the orders have started falling.

An accountant notices a change.

The originator receives a different phone call.

The relationship provides early signals.

That kind of qualitative information can be extremely valuable.

The problem is recording it without turning it into financial gossip.

There is a difference between relational intelligence and rumor.

The first is born of relevant observation.

The second is born of the human urge to fill gaps with stories.

A good professional has to know how to tell them apart.

"I heard they are doing badly" is not analysis.

"Receipts fell 30%, they lost two contracts, and they are asking suppliers for extra time" starts to be verifiable information.

That may be another role of the originator: converting signals into facts before converting them into judgment.

I do not want proximity to produce a false sense of knowledge.

I want it to produce questions that can be confirmed.

That seems a healthy balance.

Perhaps Amazon represents exactly that at scale. It does not have to depend on impressions about its sellers. It has data produced by the economic activity itself.

It is a relationship that leaves measurable traces.

Perhaps that is why more than half the small businesses financed by Amazon Lending have taken a second loan, according to the company. The relationship repeats. Each cycle generates more information about behavior.

Recurring credit can create a learning advantage.

If the first loan was paid correctly, we know something.

We do not know everything.

But we know more.

The second can be priced with a longer time series.

There is a kind of informational compounding.

That is interesting.

Most people think of compound interest only on money.

Information also compounds when relationships repeat.

Each operation adds data.

Each crisis reveals behavior.

Each payment reinforces or weakens a hypothesis.

After years, a lender can know a given client in a way a competitor cannot replicate instantly.

That creates an invisible switching cost.

The client can find a lower rate elsewhere, but the new creditor has to discover everything again.

The old creditor has memory.

If he uses that memory correctly, he can offer better terms because he knows more.

Or he can do the opposite and exploit the client's dependence by charging more precisely because he knows switching is troublesome.

Once again, the same advantage has two blades.

A relationship can reduce price by reducing risk.

It can also raise price by reducing competition.

That is why the client needs options.

An independent family office or loan broker can preserve the informational advantage of a long relationship without letting a single institution turn knowledge into capture.

That position seems more and more powerful to me.

The family office knows the family.

The banks compete.

The broker knows the company.

The lenders compete.

The main information stays close to the client, not locked inside the product.

That may be a principle I would like to follow someday.

The client should own the context.

The financial institution receives what it needs to assess the operation, but should not be the only entity able to reconstruct his economic life.

That starts to look particularly relevant at a time when data is becoming more and more important.

Whoever controls the consolidated view of the client can occupy a position similar to the one the bank historically occupied through the checking account.

Perhaps the future of the financial relationship will not be determined by who holds the money.

It may be by who best understands the person the money belongs to.

That sentence is still only a hypothesis.

But I like it.

Because it explains something I observe in family offices: wealth tends to stay loyal to the institution that knows the whole family, even when individual products are executed by third parties.

The family office does not have to manufacture every investment.

It can use managers.

Banks.

Funds.

Brokers.

Lawyers.

Loan brokers.

The advantage lies in preserving the view of the whole and coordinating the specialists.

It is a position of information before it is a position of product.

That may be why multi family offices can be especially interesting when well designed. A single family may not justify all the infrastructure needed to hold analysis, credit, investments, consolidation, planning, and a network of private operations. Several families can share part of the machine without sharing the wealth.

The risk, again, is turning the multi family office into a distributor.

If the institution has a shelf of its own that it has to sell, it starts losing independence.

Perhaps the most powerful model is the one that knows the family deeply and stays relatively agnostic about who will supply each solution.

One bank for custody.

Another for credit.

A fund for a certain asset.

A direct investment originated by another family office.

A club deal.

An off-market real estate operation.

The entire financial system becomes a supplier.

The main relationship stays with whoever organizes the table.

That is very close to what I have been looking for since the beginning of this series of letters: position.

In 2008 I began thinking that it might be more important to control the path to the buyer than to manufacture the product.

Almost a decade later, the same question appears in wealth.

Who controls the path between the family and the financial system?

If it is the bank, the family tends to see what the bank distributes.

If it is an independent adviser, the universe can be larger.

If it is the family itself through a family office, perhaps larger still.

The breadth of the map changes the options.

And options change price.

A businessman with a single lender asks whether he will be approved.

A businessman with ten qualified sources asks in which structure he should close.

Those are completely different psychological and economic positions.

Access does not guarantee good credit.

But a lack of access produces dependence.

Perhaps the most important function of a relationship is precisely reducing dependence without increasing imprudence.

Having doors.

Not having to walk through all of them.

That is power.

I come back to Amazon because this week's case condenses a good part of these ideas.

It did not begin the relationship with the merchants by granting loans.

It began by helping them sell.

The commercial activity generated information.

The information created the capacity to select.

The selection allowed offering capital.

The capital deepened the relationship.

There is a rather powerful economic circle there.

I do not know how far it will go.

There may be defaults.

There may be regulatory changes.

There may be problems nobody sees today.

It would be naive to turn a six-year-old program into a definitive theory.

But the mechanism is worth studying.

The company that controls a recurring economic relationship can end up seeing financial products before the client even approaches a bank.

That may hold for many markets.

A payments platform can advance receivables.

A marketplace can finance inventory.

An accounting software company can identify a need for capital.

A financial management system can know the company will face a mismatch before the businessman himself notices.

In that scenario, origination stops depending merely on salesmen looking for borrowers.

It can be born of the data.

That change will probably have enormous consequences for loan brokers.

The broker whose only asset is knowing a manager is threatened.

The broker whose asset is a deep relationship, interpretation, and the ability to structure may become even more valuable.

Because more data does not eliminate context.

It may even increase the amount of context required.

When everyone has information, judgment becomes the new scarcity.

That connects to last year's letter.

In 2016 I wrote that money was not scarce; conviction was.

Now I would perhaps add:

conviction improves when a relationship produces information the market does not yet have.

Not secret information.

Information built through observation.

The businessman you have known for ten years is not necessarily a better credit.

He is a credit about which you should be able to formulate better questions.

That should be the benefit.

If the relationship does not improve your ability to ask, it may be merely sociability.

And sociability should not receive a spread.

I want to carry that discipline into any activity I take up in the financial market.

If I work with businessmen, I want to know them before the need.

Not show up only on the day they need R$ 10 million.

Know the company while everything is fine.

Understand where the wealth is.

Which banks they use.

What debts they carry.

Which assets could serve as collateral.

What the financial cycle is.

Who the partners are.

What risks exist.

That way, when the demand arises, it may not be necessary to start from zero.

That comes closer and closer to a family office relationship.

Perhaps the best moment to organize credit is when nobody needs credit.

It sounds like a paradox.

It is merely planning.

When the businessman is liquid and calm, he can negotiate limits, organize documents, assess collateral, and build relationships with institutions without urgency.

When he needs the money tomorrow, his ability to negotiate has shrunk.

The good financier should work before the problem.

That holds for family liquidity too.

The time to build a line against an asset is not necessarily the day every market has closed and the client needs money.

It can be years before.

The option has value even when it is not used.

Perhaps a family office should maintain a permanent map of potential liquidity.

Cash.

Liquid assets.

Available lines.

Financeable properties.

Stakes.

Receivables.

Private sources.

Not to maximize leverage.

To maximize choice.

There is an important difference.

Debt used to increase risk unnecessarily can weaken wealth.

Access to unused debt can increase robustness.

An available line is an option.

The cost can be small compared to the value of not having to sell something at a bad moment.

That view connects credit and wealth management almost inevitably.

A family office that talks only about investments is seeing half the financial life.

The structure of liabilities matters.

Perhaps as much as the assets.

An estate of R$ 100 million with R$ 20 million of badly structured debt can be more fragile than an estate of R$ 80 million with no relevant maturities.

The photograph is not enough.

Again, context.

It is curious that this word keeps returning.

Perhaps because wealth, credit, and risk are all contextual phenomena.

R$ 10 million of debt can be enormous or irrelevant.

Collateral can be excellent or useless.

An investment can be adequate or absurd.

Everything depends on the rest.

Whoever knows the rest has an advantage.

That is why I am beginning to believe organized information may be one of the central assets of any serious financial structure.

Not merely the documents.

The memory.

The relationship.

The history.

Who decided.

Why he decided.

What happened afterward.

A loan broker should know which operations he placed at which institutions and how they performed.

A family office should know why each investment came in.

A family should be able to reconstruct its decisions even after the people who made them are no longer present.

Without that, experience disappears along with individuals.

And that may be the most underrated risk of first-generation wealth.

The founder carries a large part of the intelligence in his head.

He knows where he bought, who introduced whom, who did not pay, which bank works, which family is trustworthy, which businessman once disappointed him.

When he leaves, part of the informal family office dies with him.

The next generation receives assets.

It may not receive the map.

Turning the map into an institution may be one of the noblest functions of a true family office.

Not merely preserving money.

Preserving context.

That will probably be even more important as families increase exposure to direct investments, private credit, and club deals. The more private the portfolio, the less information will naturally be preserved by public markets.

It will have to be recorded internally.

Terms.

Documents.

Relationships.

Decisions.

Performance.

Lessons.

The wealth will become richer in information and, if badly organized, more dependent on memory.

There may be an enormous future business in organizing that.

I do not yet know the form.

For now, I observe.

But the direction is clearer and clearer.

I do not merely want to know how much someone has.

I want to understand how it behaves.

I do not merely want to look at a balance sheet.

I want to understand the economic person behind it.

I do not want to depend only on an institution to tell me whether a businessman is financeable.

I want to understand why it reached that decision.

And I do not want a relationship to work as a shortcut around analysis.

I want it to work as a source of additional information for a better analysis.

That may be the difference between knowing the client and merely having the client.

The traditional financial system is obsessed with the second expression.

"My client."

Perhaps nobody owns the client.

The client stays as long as there is a reason to stay.

The institution that knows his financial life and uses that knowledge to produce better decisions will have an advantage.

The one that uses knowledge merely to sell more products may discover that a relationship is not a prison.

It is temporary trust.

That principle holds for a bank.

For a broker.

For a family office.

For anyone who intends to work close to another person's wealth.

Knowing more increases responsibility.

If the client hands you context he would not hand to others, you receive an informational advantage and an obligation.

Using the first without respecting the second may produce revenue for a while.

It does not produce an institution.

I want to build relationships that age.

That means thinking differently about each operation.

A commission is present revenue.

The trust left over afterward is future capital.

The best deal is perhaps the one in which we obtain both.

And perhaps the greatest proof that we really know the client is being able to say no when what he wants does not match what we know about him.

An unnecessary credit.

Excessive collateral.

A club deal that increases his concentration too much.

An exclusive product that serves the ego more than the estate.

A debt whose source of payment depends on a hope.

Knowing should make the recommendation more specific.

Otherwise the relationship is decoration.

Perhaps Amazon Lending is merely one more financial product from a large technology company. Perhaps it will become much larger. I do not know.

What I do know is that the mechanism revealed something I want to keep.

The best information about a borrower may be produced long before he asks for a loan.

Credit begins before the proposal.

It begins in economic life.

Whoever can see that life clearly may hold an advantage no form filled out under urgency reproduces instantly.

That advantage does not eliminate risk.

But it may allow choosing risks others cannot understand with the same precision.

In 2015 I wrote that before credit exists someone has to find the borrower.

Now I would make a small correction.

Perhaps finding him is not enough.

One has to know him.

Leo Bentier