WHY WAS KEYNES KEEN TO INVEST IN AMERICAN BUT NOT IN BRITISH INVESTMENT TRUSTS?

Abstract

The literature on John Maynard Keynes’s activity as an investor has substantially grown in the last decade (e.g., Chambers and Dimson 2013; Accominotti and Chambers 2016; Chambers and Kabiri 2016; Cristiano, Marcuzzo, and Sanfilippo 2018; Marcuzzo and Rosselli 2018; Marcuzzo and Sanfilippo 2016, [2020] 2022). The contribution of the present paper is to investigate a specific feature of Keynes’s investment activity on his own account: his preference for American rather than British Investment Trusts. While this feature has also been observed in his investments on behalf of King’s College (Chambers and Kabiri 2016), we focus here on his personal portfolio, and we also provide a set of possible explanations for his preference. We maintain that some reasons have to do with the different structure and characteristics of the Investment Trusts in the two countries. Others relate more closely to the kind of investment policy typically adopted by the American Investment Trusts, which was much more in line with Keynes’s own approach to investment—especially regarding the stocks selection. We also attribute a role to his epistemological approach, i.e., the view that, although a full and perfect knowledge is not reachable by individuals due to the radical uncertainty characterizing the environment (“we simply don’t know,” Keynes 1937) and to the limitations of the human mind, reliable information remains, however, a guide for rational decision making, also in financial markets. Following this approach, as we will show, Keynes preferred to delegate his investment choices in the US stock market to those professionals—the managers of the Investment Trusts—who possessed, in his opinion, the wider set of reliable information on that market, while keeping for himself the investment choices in the UK stock market.

I. INTRODUCTION

John Maynard Keynes’s insights on the working of financial markets have been at the center of the so-called return to Keynes in the aftermath of the 2009 meltdown, which has seen an upsurge in the wave of references to Keynes in the media, the economic press, and political discourse. However, what the lessons are to be learned from it is far from being fully explored. There has been a lot of emphasis on the role of uncertainty in generating instability, on the importance of market irrationality, on the lack of appropriate institutional and regulatory framework, but not enough reassessment of Keynes’s views, which have remained mainly confined to his thoughts on finance from the Tract to the General Theory. While they still provide the theoretical grounds to explain them, they need to be supplemented with research on his practical acquaintance with financial markets, which guided his choices and ideas and led him to develop his own particular investment philosophy. Keynes’s notion of uncertainty remains central to his thinking, but it should not be interpreted as preventing rational decision making; on the contrary, for him, knowledge derived from information was the only means of defence against uncertainty and the guide for decision making in financial markets, whereby an investor may not always be successful but would be able to reduce the uncertainty of outcomes (see Cristiano and Marcuzzo 2018).

The literature on Keynes’s activity as an institutional investor has grown in the last decade (see Chambers and Dimson 2013; Accominotti and Chambers 2016; Chambers and Kabiri 2016) with the addition of further contributions on Keynes’s personal investment by the present authors (see Cristiano, Marcuzzo, and Sanfilippo 2018; Marcuzzo and Rosselli 2018; Marcuzzo and Sanfilippo 2016, [2020] 2022, 2022; Sanfilippo 2021), who have produced original data mainly from Keynes’s archives. Keynes’s investment approach and performances in several markets, both as a manager of institutions’ portfolios and as an individual investor, from the beginning of the 1920s until his death, have been analyzed and valuable information has been acquired, although more needs to be done to have a complete picture. We believe that his practice as an investor had an important influence on his thinking and the research on it is an essential element of scholarship on Keynes.

The contribution of the present paper is to investigate a specific feature of his investment activity: his preference for American rather than British Investment Trusts. It originates from a puzzle that emerges from the data on Keynes’s personal portfolio: his US securities portfolio was more heavily concentrated in Investment Trusts than his UK securities portfolio. This finding is not entirely new—David Chambers and Ali Kabiri (2016) had already reported this pattern when looking at Keynes’s investments on behalf of King’s College—but what is new is confirmation that the pattern is also present in Keynes’s personal investments and the possible reasons provided. His preference for American Investment Trusts is also puzzling because both contemporaries and scholars have generally assessed American Investment Trusts’ performance negatively during the 1920s and 1930s (relative to their UK counterparts). We propose three main explanations to resolve the puzzle. First, American Investment Trusts’ general management and strategies were aligned with Keynes’s own investment philosophy, which relied on the gathering of expert knowledge and information about a limited number of individual stocks rather than a large number, as instead was the case for British Investment Trusts (see Rutterford 2009; and section IV below). Second, Keynes attributed a great importance to the collection of trustworthy information to ground his investment choices, activity to which he devoted an enormous amount of time and effort, especially considering the high number of other activities in which he was engaged along his life. Although Keynes generally preferred to invest in stocks of individual companies, he did not have as much knowledge of individual US stocks as he did of British stocks and, from his perspective, investing in American Trusts was therefore a second best. In other words, in the absence of direct information on US stocks, he delegated the task of selecting US stocks to Investment Trust managers whom he considered more expert than he was in making this choice. ¹ Third, Keynes’s preference for American Investment Trusts (based on his confidence in their managers) was also related to his positive views of US businesses and entrepreneurship in general and partly inspired by Alfred Marshall.

By comparing the composition by sector, respectively, of his personal portfolios in US securities and in UK securities (see, respectively, tables 2 and 4, section III below), we give evidence of the substantial weight of American Investment Trusts in his personal portfolio in US securities (about 34% on average for the period from 1933 to 1945), as opposed to the relatively much smaller weight of the British Investment Trusts in his personal portfolio in UK securities (about 6% on average for the same period). What is relevant for our analysis is not the value of the investments in American or in British Investment Trusts in absolute terms but instead the relative weight of the Investment Trusts sector, respectively, in Keynes’s personal portfolios in US and UK securities.

In this paper we base our analysis on Keynes’s personal investments, but we provide evidence that this preference was also revealed when he was advising in his capacity as institutional investor, when he was forced to make explicit and justify his investment policy before the other members of the boards of the various institutions he managed. Clearly, there was no need to make explicit the arguments founding any investment decision or strategy when he was investing on his own. It is also not surprising that Keynes’s opinions on a single investment or sector in any given moment of time were the same when he invested for himself and on behalf of institutions.

Before addressing the question in the title, we give an overview of Keynes’s investments and the evolution of his personal portfolio over the years. Some results about his performance, based on the extant literature, are also briefly discussed, but the focus is an attempt to understand his approach and preferences about the composition of his portfolio. At the outset, it must be stressed that Keynes did not attach so much importance to diversification while he consistently followed an approach characterized by concentration, i.e., investing large units in a small set of carefully selected assets, be it shares or commodities. As already observed in the specialized literature (Moggridge 1983, in Keynes 1971–1989 [hereinafter CWK], XII, p. 10; and Marcuzzo and Sanfilippo [2020] 2022, p. 522), he preferred to bet large sums on his favorite assets (“pets”). This appears at the opposite of what we now call the “portfolio diversification approach” to risk management and can be read as consequential to his refusal of a frequentist notion of probability and his endorsement of the notion of “logical probability” (Keynes 1921, 1937).

In the following passages, Keynes clarified the rationale of his behavior: “I am quite incapable of having adequate knowledge of more than a very little range of investments. Time and opportunity do not allow more. Therefore, as the investible sums increase, the size of the unit must increase” (JMK to F. C. Scott, 6 February 1942, CWK, XII, p. 82). And again: “To carry one’s eggs in a great number of baskets, without having time or opportunity to discover how many have holes in the bottom, is the surest way of increasing risk and loss” (Memorandum for the Provincial, 7 March 1938, CWK, XII, p. 99). This naturally poses the general question of what was behind his preferences. The answer cannot but be speculative, as there may be more than one reason and cannot be ascertained for certain, but this seemed to us an exercise worth doing.

The structure of the paper is as follows. In section II we present an overview of Keynes’s investment activity and a summary of the main findings in the literature; in section III we review his investments in the London and New York stock exchanges. In section IV we investigate more specifically his investment activity on his own account in Investment Trusts, in both markets, and suggest several possible explanations for his preference for American rather than British Investment Trusts. In section V we offer some concluding remarks.

II. OVERVIEW OF KEYNES’S INVESTMENT ACTIVITY AND PHILOSOPHY

Keynes had an intense career as an institutional investor: he was director of the National Mutual Life Insurance Company in 1919, and then chairman in 1921 until 1938; director of the Independent Investment Company (1923 to 1946), the A.D. Investment Trust (1921 to 1927), and the P.R. Finance Company (1924 to 1936, chairman 1932 to 1936); and director and then chairman of the Financial Committee of the Provincial Insurance Company from 1923, remaining associated with the company in another capacity until his death. In 1921 Keynes became second bursar of King’s College, Cambridge, and then first bursar from 1924 until the end of his life.

Keynes’s investments for King’s College have been thoroughly researched, especially by Chambers in many co-authored papers. Since King’s College financial data are well organized and recorded, conclusions about Keynes’s performance as manager of King’s portfolio could be reached. According to Chambers and Elroy Dimson (2013, pp. 216, 226), “the annual performance of the Discretionary Portfolio [the part over which Keynes had full control] averaged +16.0 per cent as compared to +10.4 per cent for the market index … whilst impressive, his investment performance for King’s College was not the uninterrupted success that has previously been believed.”

To this experience as an institutional investor, there was associated an intense activity on his own account, which was a major engagement in Keynes’s life, in terms of both duration and commitment, and which he pursued among pressing public and academic obligations from his youth until his death. The Keynes Papers ² reveal hundreds of documents showing Keynes’s need and passion to keep up-to-date with the markets by constantly gathering evidence. They also show that, for Keynes, investment was a time-consuming activity in which building up a network of reliable connections and collecting sound, relevant information were costly but decisive tasks. While undoubtedly the evolution of Keynes’s economic thinking was influenced by the development of his business skills, achieved through an increase in his business contacts and sources of information, the direction of influence ran also in the opposite direction: his knowledge and method as an economist guided the selection of the sources of information and their assessment. In the case of his investment in securities, for example, factors like knowledge of the sector or company (often acquired through personal connections), the solidity of the ownership and trust in the management, together with size and frequency of dividends, may lie behind Keynes’s choice to hold on to a smaller subset of company shares more than others that, following Keynes, we have labeled his “pets”: ³

[I]t is out of … big units of the small number of securities about which one feels absolutely happy that all one’s profits are made…. I am sure I could pick up six of my pets, and that much more than the whole of our profits have been made out of them…. This has been my uniform experience in all investment connections. (JMK to F. C. Scott, chairman of the Provincial Insurance, 10 April 1940, CWK, XII, p. 78; italics added)

This is what we call his “pet approach” to investment. What we mean is that Keynes concentrated his portfolio choices on a few selected assets about which he felt very knowledgable. He referred to them also as “favourites” and “specialities,” but we have retained the more familiar terms “pet” and “pet philosophy” to label his chosen assets and portfolio approach to investment. Chambers and Kabiri (2016) use the term “core holding” and their definition is very similar to the one we employ here (see section III, fn. 8).

In his Memorandum for the Provincial (7 March 1938), he explained the difference in the performance of the three institutions whose investment he was totally or partially responsible for:

The Provincial has lost, compared with the other two, by having too small a unit of investment in particular securities. Although the funds of the Provincial are almost two and a half times those of King’s, the average unit of investment for the two institutions is about the same, the Provincial investments being about two and a half times as numerous as those of King’s. What is even more striking, although the funds of the N[ational].M[Mutual]. are nearly four times those of the Provincial, the number of the Provincial’s separate investments exceeds those of the National Mutual by about 50 (King’s 130 investments, Provincial 325, N.M. 275 in round numbers). Now it is vastly easier to find 130 satisfactory investments than to find 325; particularly if you are largely depending on the repertory of one man (the N.M. has a greater variety of sources of information). I myself follow fairly closely, or think I have some knowledge, of upwards of perhaps 200 investments; and whilst, say, 50 others (at the outside) may be followed closely by other members of the Finance Committee, I should say that the Provincial holds 50 to 100 securities about which none of the Board know much. Now out of the 200 which one tries to follow more or less, there are probably less than 50 in all classes about which, at any given time, one feels really enthusiastic. I am convinced that the good results shown by King’s are mainly due to the large proportion of its assets held in the less than 50 ‘favourite’ securities. (CWK, XII, p. 99)

As to the questions of how successful was Keynes as a short-term speculator and long-term investor, and how much money did he make, it is difficult to come up with definitive answers in the case of his personal investments. This is mainly because we lack complete information on Keynes’s personal portfolio over the entire period of his investment activities, since data are not regularly recorded and are sometimes very difficult to decipher, being scattered throughout many handwritten pages and files in Keynes’s Papers (see fn. 2), thus severely limiting the possibility to apply standard portfolio analysis. However, by looking at the composition and size of his personal portfolio, we find important clues to understand his choices and risk profile.

The recent literature has made some progress over the early estimates presented by Donald Moggridge—his tables provide only annual aggregate figures per type of assets on the total portfolio of sterling and dollar securities (see Moggridge 1983, in CWK, XII, p. 14)—so that now we have better figures for some components of Keynes’s portfolio and for some intervals of his activities. Even if these results are not sufficient to allow a conclusive assessment of his overall performance, we have a better idea of the size of his portfolio over time and some partial indication of his achievements.

Investment in currencies (from January 1920 to May 1927 and from October 1932 to March 1939) was a substantial part of his portfolio, peaking at £100,000 in 1923 in the first period and £250,000 in 1936 in the second period (see Accominotti and Chambers 2016). Investment in commodity options (from 1921 to 1939) amounted to a little less than £50,000 (Marcuzzo and Sanfilippo 2016). In commodity futures in the 1920s, his total purchases amounted to £2,362,420, while in the 1930s they amounted to £846,709 (Marcuzzo and Sanfilippo 2022). Until the early 1930s the size of Keynes’s personal portfolio in sterling securities was considerably below £50,000 (except in 1924 and 1927), to reach a peak of more than £400,000 in 1936, and it remained quite large until his death (the value of his end-of-year portfolio in 1945 was about £330,000) (Marcuzzo and Sanfilippo [2020] 2022). In Wall Street, where he started to regularly invest in 1932, the size of his dollar portfolio reached a peak in 1936, with $1,400,000 (almost £300,000 at the then-current exchange rate), but it reduced to about $250,000 (something more than £60,000) in 1939 (Cristiano, Marcuzzo, and Sanfilippo 2018) and to less than $130,000 (about £30,000) on average during the 1940s (Sanfilippo 2021).

In conclusion, the extant literature reveals that he was exceptionally gifted as a trader, not in terms of the gains he made—which, as it has now been proved (e.g., Accominotti and Chambers 2016; Marcuzzo and Sanfilippo 2016), were not as large as commonly believed—but by virtue of his deep grasp of the fundamentals underlying price trends. He showed great ability in gauging the direction of prices, although he did not always get the timing right. The literature also shows that throughout his life he was an active investor, while his own investment philosophy characterized by a “buy and hold” strategy matured only during the 1930s (e.g., Chambers and Dimson 2013; Chambers and Kabiri 2016; Cristiano, Marcuzzo, and Sanfilippo 2018). In his portfolio choices, both personal and institutional, Keynes relied on several sources, providing him with professional (as distinguished from confidential) detailed information and business analysis.

III. INVESTING IN THE LONDON STOCK EXCHANGE AND IN WALL STREET

Although Keynes had tried his hand in the London Stock Exchange (LSE) since the early 1920s, he became a serious investor in shares and bonds with his own personal portfolio only after 1932, when the LSE rebounded after the 1929 collapse. The number of companies (or local and national government issuers) in which Keynes invested was slightly over 200, with a great variety of instruments (ordinary shares, preferred, debentures, other fixed interest securities), but as far as his “valuations” are concerned (see fn. 2), we have data for only approximately 150 companies.

Keynes started to regularly register the value of his holdings (at market prices) at the end (or the beginning) of each year only from 1922 onwards, when the dimension of his investment reached a certain level (above £10,000). Therefore, our elaborations on Keynes’s UK securities holdings relate to the period from 1922 to 1945, although from other documents we know that his trading in sterling securities started well before (precisely in 1911).

Considering the whole period from 1911 to 1945, we have found evidence of 205 companies traded by Keynes (see Table 1 below, first row). Focusing on the period from 1922 to 1945 (for which Keynes’s own evaluations of the end-of-year portfolio are available), we found that fifty companies represent the pet companies, which we have defined as the companies whose shares Keynes held for more than four consecutive years (the sum of the second and the third rows in Table 1). If we take into account also a dimensional criterion (a weight in Keynes’s sterling securities portfolio greater than 1%), the number of pet companies falls to thirty-eight (second row in Table 1).

Table 1. Total number of companies traded by Keynes in sterling, 1911 to 1945

Source: Marcuzzo and Sanfilippo ([2020] 2022)

For a better understanding of Keynes’s preferences in terms of investment sectors, we have grouped the companies traded by Keynes into seven sectors: 1) Mining; 2) Industrials; 3) Transport; 4) Finance; 5) Soft raw materials; 6) Government and corporate bonds; and 7) Miscellanea, which include Utilities and also companies for which we have not found information.

For the period from 1933 to 1945 (for which we have complete data), we have computed the weight by sector in his UK securities portfolio (see Table 2 below). We also provide weight for each of the subsectors: Tin, Gold, Coal and Other Metals, and Oil (Mining); General Industries and New Technologies, the latter further subdivided in Motor and Airplanes, and Telegraph and Telephone (Industrials); Railways and Shipping (Transport); Investment Trusts and Banks and Insurance (Finance).

Table 2. Composition by sector of Keynes’s own UK securities portfolio (market value, £), as percentage of the total of each year, 1933 to 1945

* The weight of the subsector Telegraph and Telephone in the end-of-year portfolio is equal to zero for all the years here considered, except in 1934, when it was 0.7, and it has been added to the Motor and Airplane weight for that year.

^** It includes also the companies on which we have not found information.

Source: Marcuzzo and Sanfilippo ([2020] 2022)

From these data on Keynes’s investment on the LSE, the relatively small weight of the Investment Trust sector for the whole period considered clearly emerges.

After Keynes became a substantial investor in the London Stock Exchange, from 1933 onwards, we observe that, except for the year 1934, his pets accounted for about 75% or more of his end-of-year sterling securities portfolio (market value) in each year (see Table 3 below). ⁴

Table 3. Weight of Pets (%) in Keynes’s sterling securities end-of-year portfolio, 1933 to 1945

Source: Marcuzzo and Sanfilippo ([2020] 2022)

Regarding Wall Street, we know that Keynes had tried some sporadic dealings in the US companies between 1911 and 1913 (KP SE/11/1/3) and then again from 1925 to 1928. However, none of these attempts ever reached a noteworthy scale. It was during the 1930s that Keynes’s American dealings grew significantly. Until the beginning of 1934 Keynes’s inclination was that there was no immediate prospect of recovery in the US economy, and he became confident in the US market and the US economy only after the first months in office of the new Roosevelt administration. He started to invest with greater regularity in the US stock market from mid-1932 (Cristiano, Marcuzzo, and Sanfilippo 2018); trusting that the crisis of 1929 to 1931 was over, he began to purchase, for himself, the Provincial, King’s, and National Mutual assets that he thought would benefit from the recovery. He continued to buy US securities despite the banking crisis in 1933, although on a smaller scale, but he soon resumed increasingly large-scale dealings in 1934, through 1935, 1936, and early 1937. By 1938, he had liquidated a large part of his US investment, which, however, still remained in his portfolio well into 1945, while on a significant smaller scale, ⁵ the war and his involvement in the Treasury having changed the circumstances. ⁶

Two sectors stand out in his US securities portfolio: Utilities and Investment Trusts, followed by Mining and Oil. Other sectors in which Keynes invested from 1934 onwards were the motor industry and a varying set of companies in the Industrial and Consumer Goods sector, but for smaller quotas, especially from 1939 onwards. By the end of 1934, Utilities and Investment Trusts shares together already represented about a third of Keynes’s dollar securities investments (see Table 4 below). From 1937 onwards Keynes’s US portfolio did not change very much in terms of the strong preference for Utilities and Investment Trusts sectors, but it inevitably changed in value, due to the recession, the fall of the market, and large liquidation by Keynes.

Table 4. Composition by sector of Keynes’s own US securities portfolio (market value, USD) as % of the total of each year, 1933 to 1945

Source: Sanfilippo (2021)

Within seven of the eight sectors ⁷ in which Keynes invested in Wall Street over the whole period from 1933 to 1945, we find the subset of twenty-three companies, whose shares Keynes held for a period of at least four years, his American pets. ⁸

If we focus in particular on the period from 1933 to 1939, when the dimension of Keynes’s investment in the New York Stock Exchange (NYSE) was greater, we compute that his American pets accounted on average for about 75% of his end-of-year US securities portfolio, with the highest value in 1938 at about 87% and the lowest in 1934 at about 60% (see Table 5 below). ⁹

Table 5. Weight of Pets (%) in Keynes’s dollar securities end-of-year portfolio, 1933 to 1939

Source: Cristiano, Marcuzzo, and Sanfilippo (2018)

IV. INVESTMENT TRUSTS

Keynes was a director of various British Investment Trusts, the Independent Investment Company (from 1923 to his death), the A.D. Investment Trust (1921 to 1927), and the P.R. Finance Company (1924 to 1936, and chairman from 1932 to 1936; see Moggridge 1983, CWK, XII, p. 1), and not surprisingly he preferred to invest in these companies, which also represented the bulk of his pets in this sector. His activity as investor in the LSE started earlier than in Wall Street and so his holdings of British Investment Trusts cover a longer period (from 1911 to 1945). Keynes’s holdings in this sector were the following:

1. 1. A.D. Investment Trust (1922 to 1926)

2. 2. British Countries Trust (1936 to 1943)

3. 3. Burma Corporation (1924, 1927, 1929)

4. 4. Independent Investment Co. (1924 to 1945)

5. 5. London Trust (1911 to 1915)

6. 6. P.R. Finance Co. (1922 to 1935)

7. 7. Selection Trust Ltd. (1935 to 1945)

Keynes’s investment in the US stock market started only at the beginning of the 1930s, and therefore his holdings of American Investment Trusts cover only the period from 1932 to 1945. In the NYSE, Keynes’s holdings in the Investment Trusts sector were the following:

1. 1. Atlas Corporation (1935, 1936, 1938)

2. 2. Blue Ridge Corporation (1935, 1936)

3. 3. Chicago Corporation (1936 to 1945)

4. 4. General American Investors (1934 to 1940)

5. 5. Pennroad Corporation (1936 to 1940)

6. 6. Prudential Investors Corporation (1932 to 1943, not 1933)

7. 7. Selected Industries (1932, 1935)

8. 8. Tri-Continental Corporation (1934 to 1945)

9. 9. US & Foreign Securities (1932 to 1937)

10. 10. United Corporation (1938 to 1945)

Of these ten, only Atlas, ¹⁰ Blue Ridge, and Selected Industries were not pets, according to our definition, while General American Investors and Tri-Continental Corporation represented the two pets in which Keynes invested more in terms of value; Chicago Corporation, Prudential Investors, Tri-Continental again, and United Corporation were the Investment Trusts he kept longer, as they appear in his portfolio evaluations without interruption from about the mid-1930s to 1945. Except in the case of US and Foreign Securities, ¹¹ Keynes usually held common shares of these companies (see Table 6 below).

Table 6. List of US Pet securities in the Investment Trust sector held in Keynes’s own portfolio, 1932 to 1945, as of 31 December of each year

Legend: P=preferred stock; C=common stock

Source: Authors’ elaboration from Keynes’s own end-of-year evaluations (KP SE/11/5-7)

While the rationale is not clear for his choice of these Investment Trusts, we believe that it was the reputation of the founders ¹² and managers of these companies that played a relevant role.

We have found specific evidence of Keynes’s interest (at least ex post) in the performance of the US Investment Trusts that he held in his portfolio. On 15 December 1938, Buckmaster & Moore (Keynes’s broker) sent to Keynes a letter with a memorandum enclosed, which had been requested by Keynes, “giving the results of our [B&M] investigation into the experience of certain American Investment Trusts during the last eight years” (KP SE/2/07/112). ¹³ The data provided cover the period from 1930 to 1938 (up to September) and they offer a comparative picture of the performances of the following American Investment Trusts: Selected Industries, Tri-Continental, ¹⁴ Blue Ridge, Prudential Investors, US and Foreign, US and International, and General American Investors, together with data, by year, on the stock market performance, in terms of “Dow Jones % movements. Industrials.” As it is extensively explained in the accompanying note on the method of calculations, “the performance in any given year is the percentage difference between the net Assets based on the market values at the beginning of the year and the net Assets at the end of the year” (KP SE/2/7/113). The B&M employee added a caveat before concluding the letter: “I’m afraid no exact method of comparison is possible, but I hope the method adopted will show the true picture as near as possible,” and a justification in the post scriptum: “We have been unable to include Atlas Corporation owing to the complication arising out of its Subsidiary Companies” (KP SE/2/7/112). Both sentences well testify to the difficulties in acquiring data about American Investment Trusts and in comparing their performance.

Keynes found the memorandum “interesting” and a few days later, on 19 December, he sent it to Richard Kahn. Keynes added the following comment: “The superior record of General American appears practically all through.” As a matter of fact, Keynes proved right in his investment in General American since 1934, although he definitely stopped his investments in this Trust in 1940.

In the same handwritten note, Keynes also expressed some doubts about Lehman Corporation and Atlas, concluding that investment in the latter was “impracticable,” and in fact he did not invest anymore in this Investment Trust after 1938 (see also below).

On the memorandum two preliminary considerations can be made. First, the investigation/comparison provided by the broker refers to four US Investment Trusts (Tri-Continental, Prudential, US and Foreign, and General American Investors), which are among Keynes’s pets in this sector, and another two (Selected Industries and Blue Ridge) in which Keynes invested, although much more sporadically, in the years from 1932 to 1936, so it seems they have been selected primarily because they are of interest to Keynes, combined with the circumstance that data collection and analysis was possible. In fact, although Atlas Corporation is another US Investment Trust in which Keynes invested in 1935, 1936, and, to a lesser extent, also in 1938, the study did not include it because of the difficulty in collecting data, as we have seen.

Second, the comparison with the NYSE index refers not to the overall stock market but specifically to Industrials assets, therefore the focus seems to have been the comparison with the performance of the underlying assets in which American Investment Trusts were mainly invested.

According to the information provided in the memorandum, Blue Ridge, an Investment Trust in which Keynes invested very little, ¹⁵ had the worst performance in terms of the overall increase ¹⁶ in the net assets value, while General American Investors, the second Investment Trust in terms of size in which he invested during the 1930s, had the best overall performance. According to B&M investigation, Tri-Continental and Prudential, in which Keynes invested a lot and kept longer, outperformed the market in most of the years over the period from 1930 to 1936. The broker also explains that the group has shown poor records in the two years 1937 and 1938 because of some mistakes in the management. In fact we have found evidence that Keynes was disappointed by Tri-Continental “bad management,” which compared unfavorably with General American, whose management seemed to him “to have been pretty good” (it is not clear whether Keynes was referring here to portfolio management or management in general) (JMK to D. S. Roswell, 4 July 1939, KP BM/3/397). It was reported to him, by one of Keynes’s trusted advisors, that there was an organization problem, which they were going to address (D. S. Roswell to JMK, 18 July 1939, KP BM/3/401).

If we focus more specifically on Keynes’s Investment Trusts holdings in 1939 (just the year after he received this memorandum), we found that General American, Tri-Continental, and Prudential Investors shares represented, respectively, 18%, 10%, and 8% of the value of his portfolio in US securities for that year. These are similar for 1940, when the shares in these three Investment Trusts represented 38% of the value of his US securities portfolio in that year. Looking at the evolution of his US securities portfolio for the period from 1941 to 1945, we observe that the weight of Tri-Continental ¹⁷ shares remained important, 27% on average for the period from 1941 to 1944, and reached 61% in 1945 (our computations on the basis of Keynes’s own evaluations, KP SE/11/7).

We now turn to the question of Keynes’s preference for American rather than British Investment Trusts. In fact, what seems relevant to our analysis is not a comparison in terms of the dimension of the investment, accounted for by the relative weight of American and British Investment Trusts in the total portfolio in securities (sterling plus dollar), ¹⁸ but instead the comparison between the weight of the Investment Trust sector, respectively, in the dollar portfolio and in the sterling portfolio.

In both stock markets (Wall Street and the LSE), Keynes had some possible alternatives for his investment, but in the case of his dollar securities portfolio, the distribution of his investment by sector clearly witnesses a strong preference for (American) Investment Trusts, since the weight is systematically above 25% over the period considered and in many years above 30% (for a comparison with the other sectors, see also Table 4 above), while in the case of his sterling securities portfolio, the same strong preference for (British) Investment Trusts not only does not appear—since the weight of this sector is systematically below 10%—but Keynes’s choice instead went towards securities of companies belonging to other sectors, mainly Mining in the 1930s, and Transport and, to a lesser extent, Industrials in the 1940s (see also Table 2 above).

Chambers and Kabiri (2016, Table 1), showing the importance of Investment Trusts in King’s US securities holdings, average weighting being the highest (29%) component, argue that since Investment Trust stock prices can diverge from the underlying value of their investments, they can trade at either a premium or a discount to their net asset value (NAV). They write: “Although some trusts traded at discounts from 1932 onward, three of Keynes’s investment trust core holdings—Tri-Continental Common, General American Investors Common, and Prudential Investors Common—appeared to trade at a premium to NAV at those times when Keynes was an active buyer” (2016, p. 311). They also confirm that “the U.K. stock portfolio had very little if anything in investment trusts” (p. 312).

The preference for American Investment Trusts is even more striking if we consider at least two circumstances. The first one is that the American Investment Trusts, which developed in the US a bit later than in the UK, had a bad reputation (in terms of transparency and good management), especially from the end of the 1920s onwards. In fact, many were seen as corrupt and thought of as having contributed to the 1929 crash. According to a study on Investment Trusts commissioned by the Congress and undertaken by the Securities and Exchange Commission from mid-1935 to spring 1940:

During the period between the early 1920s when the investment companies first made their appearance in this country [the US] and the present [1940], approximately 1300 investment enterprises of all types were created. However, only about 650 trusts and companies are still in existence, the remainder having disappeared either through mergers, receivership, dissolution or bankruptcy. In addition, numerous companies controlled or influenced by investment companies went bankrupt or sustained substantial losses. A large portion of these losses is directly attributable to those managements which refused to recognize their fiduciary obligations to their shareholders and subordinated the interest of the investors to their own pecuniary advantage. (Release for P.M. Newspapers, 14 March 1940, Senator Wagner Introduces Investment Trust Legislation, p. 1, https://www.sechistorical.org/museum/galleries/icr/icr02_rules_of_new_game.php#ftn8, accessed February 19, 2024) ¹⁹

The second circumstance is that most of these American Investment Trusts showed a poor record (see Allen 1938; Rutterford 2009). In the same document quoted above we also read that during the 1930s, “there have been approximately 4,500,000 holders of certificates or shares of investment trusts and investment companies located in every State. American investors have sustained losses exceeding $3 billion, out of a total investment in such companies aggregating $7 billion” (Release for P.M. Newspapers, 14 March 1940, Senator Wagner Introduces Investment Trust Legislation, p. 1).

On the basis of the above-mentioned features, Keynes’s preference for American Investment Trusts appears to have been a quite hazardous choice, if these were the main elements that he took into consideration in his choice. We believe that, in accordance with his pet approach to investment, there are other explanations that may be given for his behavior; in what follows we examine separately those that seem to us to be the most relevant.

First, we consider the different structure and characteristics of the Investment Trusts in the two countries. As Rutterford (2009, p. 169) explains well and in detail, “the typical American Investment Trust was larger than its British counterpart, but held, on average, fewer securities. […] The typical British Investment Trust held over 500 stocks by 1936, compared with 81 for their American counterparts.”

In looking at the capital structure, Rutterford says that “the average UK investment trust post-World War I had 26 per cent long-term debentures, 38.1 per cent preference shares and 35.6 per cent ordinary shares. While the average capital structure of US investment trusts was less aggressive than that of their British counterparts: 40 per cent common stock and 60 per cent ‘senior’ securities, of the latter the great majority preferred stock” (Rutterford 2009, p. 170). However, the investment strategy of American Investment Trusts was more aggressive than that of the British. The typical American Investment Trust was wholly invested in domestic equities, in contrast to the British preference for fixed-interest securities, albeit international. This characteristic reflected a different management style since American Investment Trusts focused more on capital gain rather than income, speculation rather than investment, and market timing rather than long-term safer, but lower, returns. And this, in turn, reflected a different philosophy of investment, as clarified by Rutterford (2009, p. 181):

By the 1920s, Americans were happy to invest in equities and expected fund managers to seek to achieve capital gain through leverage, market timing and ‘expert’ stock selection. In the UK, retail investors preferred the security of fixed-interest securities and were content with a relatively low return in the form of income yield in return for safety through a conservative approach to reserves and an emphasis on a relatively passive investment strategy.

The investment philosophy of American Investment Trusts appears similar to Keynes’s own investment philosophy. First of all, as far the risk management is concerned, he did not manage risk by greatly diversifying his investments; rather, he preferred to concentrate his investment in a few securities (and companies) on which the amount and quality of information he believed he possessed was greater. According to our calculations, on the basis of Keynes’s own end-of-year US portfolio evaluations, about 55% on average of Keynes’s holdings were concentrated in fewer than five assets during the 1930s, and about 75% on average in fewer than four assets during the 1940s (see Sanfilippo 2021, p. 16). While on the LSE he had a direct knowledge of the business world and sufficient connections, which made him perfectly able to select the right (and few) assets to invest in, as far as Wall Street is concerned, there is no doubt that it was more difficult for him to collect and acquire an informed opinion on the company assets. This could explain why he preferred to rely on the discretionary stock picking of expert professionals, who concentrated the investment in a reduced number of assets, rather than on probability laws.

Second, we consider the definite preference for equities rather than fixed-interest securities shown by the American Trusts in comparison with their British counterparts. This is another typical feature also of Keynes’s investment activity in securities, both on the LSE and Wall Street (see Marcuzzo and Sanfilippo [2020] 2022, pp. 513–514; and Sanfilippo 2021, p. 4).

In a speech to the annual meeting of the National Mutual, on 25 January 1928, Keynes (largely before his substantial involvement in securities investment on his own account) had already clarified his negative opinion of the investment policy of the British Investment Trusts (and insurance companies), especially as compared with the American:

[I]t is extraordinarily important that we as a nation should not become […] a rentier nation depending on interest from bonds and cut off from the living enterprises of the day, where constructive things are being done and today’s wealth is being earned […]. It would be a great misfortune if we were to see others, let us say the Americans, owning the ordinary shares, in other words, equities of the new enterprises of each generation—today, for example, oil, motors, artificial silk—whilst the life offices of Great Britain were diverting the savings of their policyholders almost exclusively into […] bonds. (CWK, XII, p. 160)

From this typical characteristic of American Investment Trusts, combined with data on the greater weight of this sector in Keynes’s US securities portfolio, we conclude that as far as his investment in the US stock market went, Keynes chose to invest in the Industrials sector not directly but relying on the investment choices of the managers of the Investment Trusts, differently from in the case of his investment in the UK stock market. The reason has to do with Keynes’s approach to decision making in investment: he gave great weight to the amount of information at his disposal, based on knowledge of the company or industry, which he steadily acquired through either personal connections or factual evidence. Having less access to this information for the US stocks, he must have decided to rely on the expertise and knowledge of the managers of the Investment Trusts.

In fact, the key difference between the British and American management structure of the Investment Trusts was the role of Investment Trust managers. As well explained by Rutterford (2009, pp. 174–175):

in the US, investment trusts had managers as well as directors, with managers providing investment expertise. In the UK, directors, as well as providing respectability and conservatism, had a key management role. In Britain, it was traditional for the small number of directors to meet weekly or fort-nightly, and to take investment decisions themselves, rather than allowing the managers to do so. This policy was widespread in the investment industry in Britain with few managers and actuaries on the boards of companies until well into the 1930s.

Keynes was in favor of an informed and managerial approach to investment that was more similar to the managing style of American rather than British Investment Trusts (Rutterford 2009, pp. 176–177).

All these considerations, which can explain Keynes’s preference, do not mean that he had no criticism against certain aspects of American Investment Trusts policy, such as their “refusal […] to pick up bargains as long as they believe that it is still broadly speaking a bear market.” According to Keynes, this was “typical of credit cycling mentality.” By that he meant “selling market leaders on a falling market and buying them on a rising one after allowing for expenses and loss of interest.” This was something that “needs phenomenal skill to make much out of it” (CWK, XII, pp. 100–101). He was not entirely convinced, in certain cases, by the conduct of some of their managers. See for instance: “Personally, I hold no Lehman Corporation since I cannot believe that their alleged skill is so much greater than that of the Managers of Tri-Continental to compensate for their nearness of their quotation to their break-up” (JMK to C. H. Recknell, 24 July 1937, KP NM/1/6/43). ²⁰

We can give an additional tentative explanation for Keynes’s preference for American Investment Trusts, which is connected with his general view of the US economy and industrial organization. Keynes made his first trip to America in September 1917, to negotiate war finance; although he stayed in Washington a month (12 September to 6 October), he did not allow himself to be impressed by American society. He never visited the United States during the 1920s, despite repeated invitations (Skidelsky 1992, p. 20); however, he followed the evolution and performance of the American stock market and kept an eye on the American situation, but it was the eye of a speculator (in currencies and commodities) rather than of an investor. His short-term view may explain why he did not predict the 1929 Wall Street collapse. It was only on the second trip to America in May and June 1931 that Keynes became really acquainted with the country and its economy. The occasion for Keynes to have a first-hand knowledge of the effects of Franklin Roosevelt’s policies was during his third visit to the US, between 15 May 1934 and 8 June 1934. He stayed mainly in Chicago, New York, and Washington, meeting all sorts of people and collecting views on the American situation. For detailed information on the economy, industry, firms, or securities, he relied on published sources or personal contacts; see Chambers and Kabiri (2016).

Keynes never found it easy to understand, let alone accept, the way the Americans governed their country. This attitude remained with him during the following years, when he was an envoy for the British Government during his several visits to the US in the 1940s. There were, however, some aspects of the American industrial organization that he seemed to have admired: utility holding companies and Investment Trusts. Since he did not invest, or invested comparatively much less, in either of these sectors on the London Stock Exchange, this indicates that his pick was due to the specificity of the American business organization rather than a preference for the sectors in themselves.

Keynes was critical of the New Deal legislation to regulate Trusts because he thought it was damaging the “animal spirits” of the entrepreneurs. In his first letter to Roosevelt in 1933, Keynes warned that the process of reform could jeopardize “the confidence of the business world” (CWK, XXI, p. 290). During the recession of 1937 and 1938, while Roosevelt blamed it on scarce competition and the market power of big business, ²¹ Keynes urged him to undertake a complete reversal of approach. In a passage from his letter to Roosevelt of 1 February 1938, Keynes seems to suggest that the wildest forces of American capitalism should be harnessed for the sake of recovery rather than reduced to impotence. He wrote:

You could do anything you liked with them, if you would treat them (even the big ones), not as wolves and tigers, but as domestic animals by nature, even though they have been badly brought up and not trained as you would wish. […] If you work them into the surly, obstinate, terrified mood, of which domestic animals, wrongly handled, are so capable, the nation’s burdens will not get carried to market. (CWK, XXI, p. 438)

While these remarks, which came much after his initial involvement in the US stock exchange, do not account for his choices as an investor, they are evidence of his appreciation of the entrepreneurial spirit of American businessmen, which he thought needed to be supported.

There may be at least one further reason why Keynes was so keen about the US (and not the UK) Investment Trusts. This further reason is probably to be found in Alfred Marshall’s influence on Keynes’s formation, especially in the field of industrial economics. ²²

Marshall, who was in the US in 1875, was impressed by what he saw and learned during his visit, in particular by the “character” of the people and their entrepreneurship. He predicted a great future for this “young” country, possibly a brighter one compared with the one he could foresee for the UK. Since the times of his obituary of Marshall, Keynes showed a deep knowledge of Marshall’s analysis of American industrial organization and structure and the influence the latter played on Marshall’s thought: “His American trip made on him a great impression, which coloured all his future work. […] He used to say […] that he was enabled to expect the coming supremacy of the United States” (Keynes 1924, p. 324). ²³

As late as 1923 in Industry and Trade is the observation, “The United States has remained young very long” (Marshall 1923, p. 141). For Marshall, a country is young when the spirit of enterprise and innovation prevails on the opposite forces of habits, custom, routine, and tradition. This, in turn, may depend on several reasons, all of which are relevant insofar as they influence the prevailing psychology. Or, in Marshall’s own words, insofar as they determine the “character” of a people, and especially of its business leaders. When Keynes read Industry and Trade—which, he wrote, “represented the fruits of Marshall’s learning and ripe wisdom on a host of different matters” (Keynes 1924, p. 370)—he is likely to have appreciated the value of Marshall’s understanding of the nature of American industrial society.

While Keynes was appreciative of British management in a few companies, especially those connected with exploitation of natural resources in the Empire (see Marcuzzo and Sanfilippo [2020] 2022), he had greater faith in that form of industrial organization—public holding companies—that was prevalent in the US. His investment choices reflect his recognition of the superiority of the American business model and industrial supremacy, and the American Investment Trusts might have appealed to him for this very reason, since they were invested in domestic equities.

V. SOME CONCLUSIONS

Keynes’s investment approach showed a marked preference towards the choice of a carefully selected subset of assets that he held for several years. He relied heavily on information relative to each individual market and asset, weighing up the quality and reliability of that information through calculation of the relevant data, the advice of experts, and his own assessment of market conditions and of other participants’ opinions. He summarized his views to the chairman of the Provincial Insurance Company, F. C. Scott, on 15 August 1934: “As time goes on, I get more and more convinced that the right method in investment is to put fairly large sums into enterprises which one thinks one knows something about and in the management of which one thoroughly believes” (CWK, XII, p. 57). In the same letter Keynes well clarified the rationale of his investments concentration as opposed to the assets diversification approach: “It is a mistake to think that one limits one’s risk by spreading too much between enterprises about which one knows little and has no reason for special confidence” (CWK, XII, p. 57).

This is an example of how Keynes’s epistemological approach influenced his investment behavior. It may be recalled here that, for him, the notion of probability is not based on a frequency distribution but on deriving logical implications from given premises on the basis of information provided by the available evidence. In this process the strength of one’s own conviction—“the weight of the argument”—plays an important role, which explains why different conclusions can be drawn according to the set of beliefs of each individual (Keynes 1921). In the case of US Investment Trusts, he preferred to delegate his investment choices to their managers because he believed in their superior knowledge of the US market.

In a letter to Kahn on 5 May 1938, Keynes explained even more clearly that the investment policy he defended “assumes the ability to pick up specialities which have, on the average, prospects of rising enormously more than an index of market leaders. The discovery which I consider that I have made in the course of experience is that it is altogether unexpectedly easy to do this, and that the proportion of stumers amongst one’s ultra favourites is quite small” (CWK, XII, p. 100).

American Investment Trusts, unlike their UK counterpart, stand out among those “favourites,” for reasons related to their typical capital structure and investment philosophy (much more similar to Keynes’s own) but also because they represented for Keynes a better alternative to his insufficient knowledge of the US market. Perhaps also the idea Keynes derived from Marshall that the performance of an industrial system depends much on the “character” of its business leaders played a part in Keynes’s preference, and it may have persuaded him that the American managers of the Investment Trusts had a more innovative approach to portfolio management than their British counterparts.