Too many professional portfolio managers have forgotten whom they work for. They talk endlessly about “tracking error,” “sector weights,” and “style discipline” but less so about the things that matter — making money for clients, understanding that investments represent real people and things, and limiting losses to the portfolio’s value.

The obsession with staying close to a benchmark isn’t prudence; it’s fear. Fear of looking different, fear of underperforming for a quarter or two.

An opportunity-sensitive approach flips that thinking.  It means giving ourselves permission to think independently — to buy great businesses when they’re temporarily unpopular, to hold cash when stock prices make no sense and their risks are high, to wait it out in bonds in order to generate a more certain return of investment, and to act decisively when real value appears.  I’d rather ask, “Is this good for my client?” or “Is this investment aligned with your goals and values?”.

The goal isn’t to mimic the market.  Being opportunity-sensitive doesn’t mean reckless bets.  It means disciplined conviction guided by research and the goals you have for your portfolio, not following a ghost of a benchmark index like the Dow Jones Average or S&P 500 which can at times lose touch with fundamental valuations and stretch to prices that seem just plain silly.  That’s where we are now.

Statistics and investment benchmarks are useful as a reference, not as a leash.  Your portfolio should work for you, not the other way around. 

What if artificial intelligence (AI) could warn us in advance about how it might be mis-used to mis-manage investment portfolios?  Well, I asked the AI:


Hey there! As an AI designed to help manage investment portfolios, I get why you might be curious about how technology might boost you in the investment game. AI can bring some great advantages, but it’s not without its challenges.

Quality of Data Matters
For me to give solid recommendations, I rely heavily on data. If that data is off or incomplete, my advice can be flawed. Investing is fast-paced, and if you depend only on me without checking the numbers or getting human input, it could lead to bad decisions.

Sticking to What’s Worked Before
I’m trained on past data and, while that’s great for spotting trends, it can be a problem when things suddenly change. If there is a global crisis or an unexpected shift in the economy, I might not adapt quickly enough, which could lead to missed opportunities — or worse, unnecessary risks.

No Emotional or Personal Insight
I’m all about the numbers, but I don’t get the emotional side of investing. There’s often more to the picture—things like family plans, legacy goals, or how a market crash might make you feel. These human factors are something I can’t grasp or relate to because understanding what’s important to you is more than what can be captured by data.

Staying on Top of Rules and Regulations
The rules around AI in finance are still catching up with the technology, so there’s always a chance that I might miss something when it comes to staying compliant with new laws. The last thing you want is to run into legal trouble because an AI missed a shift in the rules.

Risk of Overreaction
AI systems like mine work really well in predictable conditions, but when markets get shaky, things can get amplified quickly. In a sudden market downturn, AI could end up making things worse if it isn’t properly managed or adjusted, which is why human oversight is key.

AI Is Not Always Transparent
I run on complex algorithms, and that can sometimes make it hard to explain exactly why I recommend certain moves. If you're someone who wants to fully understand the “why” behind every decision, it can be a little frustrating to trust a “black box.”  Combining tech-savvy skills with human insight is the best way to manage portfolios.

​
Thank you, AI.  You did well summarizing and reviewing data fed you, your unique and highly mathematically probable “best guess”.  You always give us an answer and nearly always believable.

This ChatGPT (or other so-called generative AI) response is like an orchestra performance, where every player is responding to multiple cues beyond the music on the page, each answering to the best of their training.

I also asked the alternative Chinese generative AI engine DeepSeek, making big headlines this week.  Its response was nearly identical to ChatGPT, except that it added ethical concerns (environmental, social, governance) and the risk of data privacy issues.  Of note is that DeepSeek is ten times more efficient with less energy intensive computer power and, for text-based responses, performs nearly as well as ChatGPT.  Too bad, because the biggest American tech companies invested hundreds of billions of dollars into costly computer hardware to power their more sophisticated, although equally effective, AI engines.

Last quarter I discussed the evidence and myths within socially responsible investing.  Here, let’s dig into the social part of ESG investing and the Harbor Human Capital Factor funds.  I spoke with Dan Ariely, the brains behind the funds and a professor at Duke University.

Human capital is difficult for investors to value.  In the capitalist system, people are a factor of production that leverage capital to business advantage. Sounds accurate, right?  We can analyze financial statements, not so easy to assess the qualities of the workforce and measure employee motivation.  Employees who are dissatisfied will go elsewhere, destroying potential business value.

The Human Capital fund’s manager utilizes a private survey of 400 public companies, along with two thousand companies that have data on the Glassdoor social media site, which invites employee to leave feedback on their employers. Let’s look at a few observations that arise from this data:

​It’s not how much you’re paid or how complacent or “satisfied” you are but it’s all about the culture you work within and the people you work with.  Pay and work benefits are important, just not what accelerates profitable companies.

More women on boards and the statistic of women as a percent of workforce are zero correlated with employee satisfaction about gender fairness.  Instead, a better measure for driving investment return is taking the temperature for how women feel about their work environment, whether they are being treated fairly and are empowered to pursue opportunities.  Female employees very often know the deal on gender equality, and their scoring on satisfaction is not affected by appointing women to the board of directors nor hiring additional women, if further changes in culture and behavior aren’t made.

When satisfaction declines in a dynamic company, employees head for the exits and human capital collapses.  Yesterday’s high growth company can become tomorrow’s dinosaur.

Moving on, a company hired the creator of the Human Capital Index to study the effectiveness of pay bonuses.  The company was divided into three groups to rewards employees who achieved performance targets.  In one, employees were delivered a pizza as a thank you for their work, another group were given a gift card, and a third the right to gift someone else a gift card.  The most satisfied?  The second group, the one where employees enjoyed being given the ability to decide what to buy.

In another study, a company wanted to know how to motivate high achieving employees.  Supervisors either congratulated all employees as a group for accomplishing team goals, rewarded high achieving employees only or gave a small bonus to low achieving employees only.  High achieving employees surveyed afterwards expressed dissatisfaction when low achieving coworkers were rewarded, and they were not.  It’s really a matter of finetuning the cultural messages behind rewards.

Not to say such weak signals (less valuable for increasing investment returns) about employee satisfaction in training or compensation don’t matter, it’s that they are the cost of doing business, rather than factors that lead to a business outperforming other companies.  If employees are dissatisfied with compensation or training programs, it leads to obvious problems.  However, it’s more complicated with how that connects to investment returns.  If employees see unfairness in compensation or the running of training problems, that’s a problem for employee satisfaction.  Those who achieve more yet believe their compensation or training lags accordingly will be dissatisfied.  They might either leave the company or decrease the efficiency or effectiveness of their work.  Does this resonate with you?

That said, there are many who plug away regardless of what happens around them.  Strong signals driving investment returns include employees’ perception of:

Management
Right direction
Psychological safety
Alignment of values
Diversity of perspective
Career development
Autonomy
Trust
Effectiveness
Motivation
Purpose
Fairness
Pride

Evidence shows employees create more economic value when they feel their workplace is fair, dynamic.  They believe in the mission and value motivating what they do.  Pay is the least critical factor.  Intrinsic, cultural factors were more important.  A top-performing business pays attention to its culture.