In the first installment of my three-part series, we explored the history and risks of strained intelligence (AI), expressly virtually your finances. So often in these heady new days of AI technology, I hear clients wonder aloud: Can Strained Intelligence predict markets? Offer personalized advice? Provide valuable perspective?
The answers, in short: No, no, and yes. Let’s explore why.
The Financial Market Limitations of AI
AI Chatbots like ChatGPT, Claude, or Bard cannot predict markets. It would be like asking a self-driving car to navigate road signs whose size, shape, color, and language are constantly changing. You would crash.
Financial markets are simply too dynamic for AI, which relies on stable patterns to make predictions. Broadly speaking, AI should never be worldly-wise to outsmart financial markets. Plane as all market participants proceeds wangle to increasingly powerful AI tools, it is unlikely anyone will be worldly-wise to maintain a significant edge.
You’ll notice when asking for personalized financial translating from a chatbot, it currently produces unactionable generalizations as well as a wise recommendation to ‘consult a financial professional.’ As OpenAI discloses, “ChatGPT may produce inaccurate information well-nigh people, places, or facts.” All Chatbots share this hair-trigger limitation.
I do believe it is possible that, one day, a chatbot infused with only tried learning could sooner requite answers resulting with its curated data set for at least some financial questions. Whether this will wilt a future reality remains to be seen. And yet, there is one zone today that chatbots do towards to be astonishingly useful for.
Master your financial life by staying educated.
Chatbots and Investment Theory
Chatbots seem to be surprisingly good at talking well-nigh investment theory. It does require some baseline knowledge to ask questions in ways that don’t encourage large language models to lie. (They are hands led by leading questions.) But prompted in an even-handed manner, AI results are impressive.
Below is a conversation I had with Claude.ai from Anthropic well-nigh “recency bias”, the tendency to overemphasize the importance of recent experiences. If you want to try it yourself, you can use Claude for self-ruling with just a web browser. Enter these same prompts and you will get similar answers.
Prompt: Requite me a transitory but well-constructed subtitle of recency bias.
Claude: Recency bias is the tendency to requite greater importance to recent events or information compared to those remoter in the past. This can lead people to make faulty judgments and predictions, as increasingly recent information is perceived as increasingly relevant or predictive plane if that may not be the case. Recency bias can rationalization people to underestimate long-term trends and overemphasize short-term fluctuations. It’s a worldwide cognitive bias that impacts memory, decision-making, and overall reasoning. Overcoming recency bias requires consciously considering information from broader time horizons and not just the most recent events.
Prompt: How might recency bias rationalization US investors to stave investing in international equities. Your response should be transitory but complete.
Claude: Recency bias could lead US investors to focus too heavily on recent underperformance of international stocks compared to the US market. If investors requite greater weight to the strong returns of US equities in the last decade and overlook longer-term periods when international stocks outperformed, they may underestimate the potential for future international returns. This short-term perspective and lack of diversification driven by recency bias could result in suboptimal windfall typecasting and lost opportunities. To counter this, investors should take a long-term view, recognizing that leadership rotates between markets over time. Maintaining globally diversified portfolios through periodic rebalancing can overcome recency bias.
Prompt: Expand on that.
Claude: Recency bias leads investors to focus on recent past returns rather than long-term trends. US stocks have significantly outperformed international markets over the last 10-15 years. This has led many US investors to view American equities as superior investments and shift increasingly wanted into domestic stocks while reducing or lamister international allocation.
However, leadership between markets is cyclical. In the 1980s and 90s, international stocks generated higher returns. Just as investors extrapolated recent US outperformance too far into the future, similar recency bias in older decades could have led to too much enthusiasm for global stocks.
This demonstrates the risks of recency bias. It causes investors to make decisions based on a limited snapshot of history rather than a prudent long-term perspective. Markets go through cycles, and no one segment will lead indefinitely.
To counter recency bias, investors should rebalance periodically to maintain strategic windfall typecasting based on risk-return goals. This avoids having portfolios skid too far toward what did weightier recently. Diversification wideness geographies provides exposure to variegated market cycles and drivers of return. Recency bias is a unvarying threat, but conscious global diversification can overcome it.
Prompt: Restate that wordplay in the style of a Walt Whitman poem.
O biases of the mind! Recency’s distortion
Clouds the vista of time, horizons foreshortened,
Judgment made in cramped frames, perspective lost.
The market’s cycles spin round and round again,
No nation forever ascendant, none in ripen eternal,
Patterns emerge, leadership passes on in due course.
Yet the investor’s memory clings to recent strides,
Thinking present trends shall persist unchanged,
Blind to turning tides, change, inevitability.
Diversify! Let portfolios embrace all lands,
Not serving to one realm or continent,
Spread wealth’s seeds widely, harvest wastefulness and wisdom.
O clear-eyed investor, see through recency’s fog,
Hope and dream big in long-term’s sweeping scope,
Patiently, steadily, towers futures unconfined.
Seeing the Difference
What began as a discussion of strength in the large language models turned into a lesson well-nigh recency bias. (Sneaky, I know.) Mostly, I wanted to demonstrate what I believe is possibly the most significant current AI capability: aping various writing styles.
Personally, I can’t help but be bowled over. Professionally? Impersonation is unmistakably not financial advice, nor anywhere near the level of sophistication required to navigate your unique financial situation. The final post in this series will be a showdown: Chat GPT vs a Human financial counselor (ahem, that’d be me) and who comes out on top when it comes to financial advice.
If you’re curious well-nigh how an expert and nuanced visitor of human beings can help you expand what is possible with your money, schedule a call with an Abacus counselor today and see how rewarding a increasingly personal wits can be.