Behind Closed Doors: The Secret Life of Family Offices and the Wealth They Protect
A family office is like a superhero, except instead of saving the world, they’re saving your portfolio… one tax solution at a time ~ ChatGPT
Some of the more interesting readings I am digesting last month are regarding Family Offices. This structure is favoured from China Chinese clients as well as HNW families whom chose Singapore as a premium location to park their assets, and it is a sizable growth market for finance professionals in the long run. Singapore family offices are favored by wealthy people from all over the world because of their tax-free policy, stable political and business environment, strict financial supervision and mature trust practices and legal support!
1. What is a family trust?
Family trust is a property management method in which a trust institution is entrusted by an individual or family to manage and dispose of family property on their behalf to achieve the wealth planning and inheritance goals of the wealthy. In a family trust, the ownership of assets and the right to income are separated. Once a rich person entrusts the assets to a trust company, the ownership of the assets no longer belongs to him, but the corresponding income is still distributed to the beneficiaries (usually family members) according to his wishes. There is no need to pay inheritance tax, and because the wealth has been isolated. If the wealthy divide the family property through divorce, bankruptcy, debt collection, accidental death, etc., the money will exist independently, will not be affected, and will not be pursued. Family trusts can better help high-net-worth individuals plan for "wealth inheritance", achieve wealth for more than three generations, and pass wealth from generation to generation. This crucial area is gradually recognized by China's wealthy.
The inheritance of family wealth has always been a worldwide problem and a topic that the rich are always concerned about. In China, the 30 years of reform and opening up have created a large number of wealthy families. However, many of the business empires they built through hard work and the family wealth accumulated have not escaped the curse of "not being able to get rich for more than three generations." In contrast, in European and American countries, many well-known families, such as the Rockefeller family and the Kennedy family, have relied on the wealth inheritance mechanism of family trusts and family foundations to achieve everlasting family wealth.
Functions of family trusts
1. Asset isolation and preservation——Safety, bankruptcy protection
2. Legal tax avoidance——No inheritance tax, etc., other tax arrangement strategies
3. Freedom of contract and privacy protection
4. Reasonable and consensual wealth inheritance
a. Prevent future generations from squandering money and wasting money.
b. Prevent intra-family disputes over property
c. Eliminate the long procedures and expensive fees of probate and avoid stamp duty.
d. Family management: Establish reward mechanism, good family contract and culture
e. Wealth requires more than three generations, business inheritance, wealth inheritance, and lasting foundation.
Advantages of Singapore Family Trusts
1. Stable political and economic environment
2. Transparent and complete legal system
3. Mature operation and supervision system (trusts under the Anglo-American legal system have a history of 700 years)
4. Recognition by international financial markets
5. Strict confidentiality
6. Permanence (passed from generation to generation, and continues to increase in value)
7.Tax benefits
a. No capital gains tax, no tax on trust financial asset income
b. No tax is payable when trust assets are distributed to beneficiaries.
c. No inheritance tax or gift tax
Singapore Family Trust vs. American Family Trust
There are tax advantages for family trusts in Singapore (whether it is a foreign settlor trust or a domestic trust)
a. No capital gains tax, no tax on trust financial asset income
b. No tax is required when trust assets are distributed to beneficiaries (if the beneficiary is an American or a US tax resident, tax is required (in this case a foreign grantor trust should be established))
c. No inheritance tax or gift tax
d. There is no tax on income earned abroad when it enters a Singapore trust.
U.S. family trusts are taxable (whether foreign settlor trusts or domestic trusts)
a. There is capital gains tax, and income from trust financial assets must be taxed
b. The distribution of trust assets to beneficiaries is subject to tax
c. There is inheritance tax and gift tax
d. Income from abroad must also be taxed when it enters a trust (the United States has global taxation and super strong enforcement)
Note: The above situation is only true under certain conditions. The specific conditions depend on the relevant terms.
How to set up a Singapore family trust
(1) Consultation and design are comprehensive. Family trust involves many aspects such as law, financial management, taxation, etc.
(2) List assets to be transferred to the trust (list assets)
(3) State the purpose of establishing the family trust (state goals)
(4) Choose a trust company (Trustee)
(5) List Beneficiaries
(6) Establish a protector to supervise
(7) Designated investment adviser (Financial adviser)
Although I am not at the right capacity to start a family office at the moment. Some of my friends and clients whom are more financially successful than me has probed me about it. I plan to research more on this so that eventually, I will be able to utilise this tool and recommend the required services and solutions to the right person when the time is right.
Portfolio Decision for the month of Jan.
January is a bloodbath for the equity markets. Deepseek upended the perceived US monopoly of Large Language Models by US Big Tech, and created a low cost model whom can perform at considerable specs for a fraction of ChatGPT training cost. Liang Wenfeng (Hedge fund manager of quant fund High Flyer) created a side project whereby he acquired 10,000 Nvidia A100 GPUs before the United States restricted chip sales to China in 2021. His team trained their R1 model for US$6 million compared to $100 million for OpenAI's GPT-4 in 2023, and approximately one tenth of the computing power used for Meta's comparable model. Moreover, Deepseek is an open source weights model whereby any single computer user can simply copy the source code and deploy their own LLM on their own terminals. Upon the release of the news, Nvidia's share price dropped by 18% on a single day and investors are flooded by the Sea of Red at Chinese New Year.
Reviews on Deepseek are mixed. Within the technical crew that I discuss technology issues on, common criticism are the possible overstatement of performance metrics and understateement of input cost on Deepseek. Deepseek is running on highly subsidised energy cost and managed to pull off a miracle and procure impressive number of chips that were considered impossible to get. For training data wise, Deepseek used alternative methods whereby they pool logic and knowledge using a (Mixture of Experts) Methodology and use multiple LLMs to train their own AI. This resulted in Deepseek unable to operate independently on their low cost methodology without the underlying base LLM.
Based on my own user testing, Deepseek lag considerably in depth and detail on current affairs and data for publically avaliable (but technical) queries. They are hit by huge lag and outages as they did not build up their servers and capacity before launching their product to the entire world. This means their reported base cost will be drastically different pre launch compared to post launch if they wish to scale the usage up. Also, censorship issues and data privacy is a major issue enclosed within the blackbox of Deepseek, whereby the data is stored in China Chinese Servers. All these makes it unpopular for incumbent user, especially if a new solution is not 10x better than the previous one.
Essentially, based on my own understanding, DeepSeek has permenantly disrupted the oligopoly of Big Tech on high specifications LLM, and allow every company to be able to build and run their low cost LLM based on their idiosyncratic company data. The analogy will be the tectonic shift from Centralised Mainframe computers created by IBM in the 1950s, to decentralised Personal Computers led by Mac and Microsoft after that. The TAM of demand of CPU and GPU Chips will be set to increase over the long run, but the stablity of demand of High specs chips are highly uncertain. The energy demands of the US-China Tech war will be highly uncertain too as the energy requirements for Deepseek is drastically lower than the high spec LLMs and there might be shifts in the spending capex of the CTOs in big tech.
Following the above insights, I did not make any divestment or increase in my position for Nvidea and AMD. Although they are more reasonably priced now, I do not have a precise valuation given the multiple moving parts in order to determine if there is a margin of safety if I load up on these company right now.
I divested my Ovintiv position on 04Feb 25 at USD 57.0996 at a profit of 38pct . The original investment thesis of Ovintiv was the ability of US to tap on the energy and oil field of Canada to fund their energy requirements and the geographically proximity to simply build pipelines to redirect the energy flow and unlock the deep value in Ovintiv. Since Trump proposed (and retracted back to deferred sanctions) on Canada, the Trump - Canada trading relationship is permenantly imparied for the tenure of Trump’s term, rendering the addressable market moot. Moreover, the catalyst for heightened energy demand of AI / US-CN tech war is now highly uncertain. I decided to build up more cash in case there are better opportunities out there.
Going into my new role is a massive learning curve. I am learning about needs based planning philosophy as well as prospecting and comparing of the best solutions across the different investment and insurance providers. I am glad I have supportive and enthusiastic new teammates and leaders whom are willing to share and learn with me, and I hope this new year will be a great leap forward!